Robot-Assisted Responsive Neurostimulator System Placement in Medically Intractable Epilepsy: Instrumentation and Technique

Robot-Assisted Responsive Neurostimulator System Placement in Medically Intractable Epilepsy:... Abstract BACKGROUND The management of medically refractory epilepsy patients who are not surgical candidates has remained challenging. Closed loop—or responsive—neurostimulation (RNS) is now an established therapy for the treatment of epilepsy with specific indications. The RNS® system (NeuroPace Inc, Mountainview, California) has recently been shown to be effective in reducing the seizure frequency of partial onset seizures. The electrode design consists of either intracerebral depth electrodes or subdural strip electrodes, and stereotaxis is typically used to guide placement into the EZ. Details on the operative techniques used to place these electrodes have been lacking. OBJECTIVE To address the advantage of using a robotic-assisted technique to place depth electrodes for RNS® system placement compared to the typical frame-based or frameless stereotactic systems. METHODS We retrospectively reviewed our single center, technical operative experience with RNS® system placement using robotic assistance from 2014 to 2016 via chart review. RESULTS Twelve patients underwent RNS® system placement using robotic assistance. Mean operative time was 121 min for a median of 2 depth electrodes with mean deviation from intended target of ∼3 mm in x, y, and z planes. Two patients developed wound infections, 1 of whom was reimplanted. Seizures were reduced by ∼40% at 2 yr, similar to the results seen in the open label portion of the pivotal RNS trial. CONCLUSION Robotic-assisted stereotaxis can be used to provide a stable and accurate stereotactic platform for insertion of intracerebral RNS electrodes, representing a safe, efficient and accurate procedure. Robotics, Drug-resistant epilepsy, Responsive neurostimulation, Depth electrodes, Surgical technique ABBREVIATIONS ABBREVIATIONS CT computed tomography MEG magnetoencephalography MRI magnetic resonance imaging PET positron emission tomography RNS responsive neurostimulation SEEG stereotactic electroencephalography SPECT single-photon emission computed tomography Despite the relative abundance of publications describing both the concept of closed loop stimulation and the clinical efficacy of responsive neurostimulation (RNS; NeuroPace Inc, Mountainview, California) in treating patients with partial onset focal epilepsy, there are few descriptions of the surgical technique and instrumentation necessary to properly place the depth and/or cortical strip electrodes. Surgeons typically use a frame-based system to place the leads,1,2 though frameless systems have been described.3 While popular in many other surgical subspecialties,4 robotic platforms have typically had limited applicability in most neurosurgical procedures. However, robotic-assisted placement of multiple depth electrodes for stereotactic electroencephalography (SEEG) has become increasingly utilized due to the technique's ease of use, stereotactic accuracy, and efficiency.5,6 Because many patients undergoing RNS therapy require 2 depth electrode placements, we have begun to use a robotic platform (ROSA Surgical Robot, Zimmer Biomet, Warsaw, Indiana) to place these electrodes and have found it to be just as accurate as frame-based placement and more efficient in terms of operating time. This report details our experience describing the technique of depth electrode placement using a robotic platform for RNS therapy. Table 1. Patient Characteristics Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  ATL, anterior temporal lobectomy; DTI, diffusion tensor imaging. View Large Table 1. Patient Characteristics Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  ATL, anterior temporal lobectomy; DTI, diffusion tensor imaging. View Large METHODS Patient Selection A retrospective chart review was performed in order to identify patients who had undergone robotic-assisted placement of RNS (NeuroPace Inc) therapy. This study was conducted with the approval of the Cleveland Clinic Foundation Institutional Review Board, and a waiver of informed consent was approved to allow access to protected health information by the research team. In order to be considered for RNS implantation, patients must have medically refractory simple partial or complex partial seizures, with or without secondary generalization. In order to localize the epileptic focus, our institution uses a variety of methods in the presurgical evaluation including video electroencephalography, magnetic resonance imaging (MRI), ictal single-photon emission computed tomography (SPECT), positron emission tomography (PET), and magnetoencephalography (MEG; Table 1). If the noninvasive information is discordant, invasive evaluation with SEEG or subdural grids and depth electrodes may be utilized. After this evaluation, candidates for RNS typically demonstrate bilateral mesial temporal ictal onset, unilateral mesial temporal onset with inability or unwillingness to tolerate verbal memory deficit or a focus in eloquent cortex (Table 1). Surgical Procedure The operative trajectories for depth electrode placement are planned on the robotic platform's surgical navigation software before the procedure (Figures 1A, 1C, and 2A). The patient receives prophylactic antibiotics and is placed under general anesthesia with full muscle relaxation for the entire procedure as once the patient's head position is stereotactically registered to the robot, the patient's position cannot change. For hippocampal depth electrodes, we typically place the patient in the prone position. The patient's head is fixed to the table with a standard 3-point Mayfield clamp. The robot is then registered to the head either via facial scanning using software included with the robot or using scalp fiducials if prone. Six facial landmarks are used for markerless stereotactic registration (left and right medial canthi, and lateral canthi, glabella, and nasal tip). The error is then calculated with a corrected offset value in each of the registration points. Once this error is noted to be 1 mm or less, registration is then validated and adjusted and the surgeon scans 4 facial areas (bilateral temporal and nasal areas) before the final registration verification. FIGURE 1. View largeDownload slide Robotic-assisted placement of responsive neurostimulator device. A, Illustration of surgeon planning operative trajectory on software preoperatively and hippocampal electrode placement. Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography ©2014-2017. All rights reserved. B, Preoperative MRI-planned trajectory of a hippocampal electrode used for responsive neurostimulation. C, Postoperative CT demonstrating accurate hippocampal electrode placement. FIGURE 1. View largeDownload slide Robotic-assisted placement of responsive neurostimulator device. A, Illustration of surgeon planning operative trajectory on software preoperatively and hippocampal electrode placement. Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography ©2014-2017. All rights reserved. B, Preoperative MRI-planned trajectory of a hippocampal electrode used for responsive neurostimulation. C, Postoperative CT demonstrating accurate hippocampal electrode placement. FIGURE 2. View largeDownload slide View largeDownload slide Bilateral hippocampal electrode placement for responsive neurostimulation. A, Bilateral hippocampal preoperative planning trajectories. B, Postoperative CT demonstrates accurate placement bilaterally. C, Blended CT/MRI demonstrating accurate placement bilaterally. FIGURE 2. View largeDownload slide View largeDownload slide Bilateral hippocampal electrode placement for responsive neurostimulation. A, Bilateral hippocampal preoperative planning trajectories. B, Postoperative CT demonstrates accurate placement bilaterally. C, Blended CT/MRI demonstrating accurate placement bilaterally. The surgical incisions are planned at both the electrode insertion sites and for the pulse generator. If the patient has had prior surgery, the old incision will need to be incorporated into the new incision or re-used, using the pulse generator template in order to plan an incision large enough for placement. If the patient has not had prior surgery, typically a barn door or “U” shaped incision is chosen for the pulse generator while the electrode insertion site incisions are variable depending on the electrode location. The surgical region is prepped and draped in the usual, sterile fashion. The robot is then draped under a sterile, plastic cover. A guide tube with a 2.5 mm inner cannula is attached to the robot arm. The desired trajectory is selected on the touch screen interface. After trajectory confirmation, robot arm movement is initiated through the use of a foot pedal. The robot arm locks the drilling platform into a stable position once reaching the calculated trajectory. The distance between the target site and the top of the guide tube attached to the robotic arm is pre-set at 150 mm. At the entry site, a 2.5 mm diameter hand drill bit (Stryker, Kalamazoo, Michigan) is used for making an incision and trephination in the skull. The dura is coagulated using monopolar cautery and gently punctured using a thin obturator probe. The robotic arm is then moved close to the skull to calculate a distance to the target. The electrode depth is marked and inserted. At this point, care is taken to avoid lead migration during stylet removal. A silicone cap device (NeuroPace burr hole cover model 8110; NeuroPace Inc, Mountainview, California) is placed on the bone, the plastic sheath sleeve is placed covering the electrode and secured to the bone with a titanium plate and screws. After electrode placement is completed, the robotic arm is moved out of the operative field and the process repeated for all trajectories. To place the pulse generator, the marked incision is made as outlined above. Frequently, temporalis muscle will have to be elevated in order to access the outlined region of skull. The bone is then marked out according to the dimensions of the provided ferrule, a burr hole drilled, and a craniectomy performed in the shape of the ferrule template. Care is taken to avoid penetrating through the inner table of skull, thereby potentially tearing dura or underlying cortical vessels. Bone and epidural bleeding is meticulously controlled with a combination of bonewax and hemostatic products. The provided ferrule is implanted and secured to the skull with 4 self-tapping screws (1.5-mm diameter, 4 mm long). The tunneling tool is passed such that the electrode leads are tunneled to the region of the pulse generator. The pulse generator is brought into the operative field, connected to the distal end of the depth electrodes, and secured in the ferrule. If there are additional electrodes that may be used in the future, a subgaleal pocket can be made adjacent to the burr hole site. It is important not to create a pocket near the pulse generator in order to avoid lead damage during battery replacement. The programmer wand is covered with a sterile bag to interrogate the pulse generator. The lead impedance and real-time electrocorticography are verified to ensure the leads are recording properly. The surgical wound is thoroughly irrigated and closed in anatomical layers. Dressings and a head wrap are applied. Postoperatively, a volumetric computed tomography (CT) scan (1-mm cuts) and skull X rays (anteroposterior and lateral) are obtained in the recovery room before sending the patient to the surgical floor (Figures 1B, 1D, 2B, 2C, 3B, and 3C). FIGURE 3. View largeDownload slide Placement of responsive neurostimulator electrodes in eloquent cortex. A, Preoperative MRI demonstrates prior resection just anterior to primary motor cortex. Pathology was focal cortical dysplasia and patient's seizures recurred. SEEG localized the focus to the margins of the resection cavity, and therefore responsive neurostimulator electrodes were placed around the margins of the resection, including in primary motor cortex. B and C, Postoperative skull x-ray and CT demonstrating lead placement. FIGURE 3. View largeDownload slide Placement of responsive neurostimulator electrodes in eloquent cortex. A, Preoperative MRI demonstrates prior resection just anterior to primary motor cortex. Pathology was focal cortical dysplasia and patient's seizures recurred. SEEG localized the focus to the margins of the resection cavity, and therefore responsive neurostimulator electrodes were placed around the margins of the resection, including in primary motor cortex. B and C, Postoperative skull x-ray and CT demonstrating lead placement. Data Collection and Statistical Analysis Demographic, operative, and clinical outcome data were collected from review of patients’ electronic medical records, and descriptive statistics were applied in Excel (Microsoft, Redmond, Washington). “Prep time” was considered to be the time marked as “ready for surgeon” to skin incision. “Operative time” was considered to be from skin incision until “procedure end” which is typically marked when the dressings are applied to the head. Data on the stereotactic accuracy of the procedure were collected using the robotic software (Rosana, Zimmer Biomet, Warsaw, Indiana). A postoperative CT was imported to the software and fused to the planning MRI and CT angiography. Target location errors were then calculated by measuring the distance in each plane (x, y, z) from the center of the trajectory targets and the center of the actual distal electrode contact. RESULTS In our institute from 2014 to 2016, 12 patients with localized focal epilepsy were successfully implanted with the RNS® (NeuroPace Inc) system using robotic assistance. Patient characteristics are listed in Table 1. Patients had a diagnosis of epilepsy for an average of 21 yr. Every patient had some type of prior surgery in order to attempt to localize the epileptogenic zone (either SEEG or subdural grids/strips) and 4 patients had a prior lobectomy which ultimately failed. The majority of patients had bilateral mesial temporal seizure onset zones as determined by either SEEG or depth electrodes. Table 2 demonstrates operative data and postoperative outcomes. The procedure was performed in an average operative time of 121 min (range 70-184 min) with 65 min of prep time (range 42-98 min) for a median of 2 electrodes per surgery. Twenty-six total electrodes were placed. The mean absolute error for electrode placement was 3.2 ± 0.8 mm in the x plane, 3.4 ± 0.6 mm in the y plane, and 2.9 ± 0.6 mm in the z plane such that the mean Euclidean error was 5.5 ± 1.1 mm. All patients did well immediately postoperatively without any neurological decline or other complications. The median hospital stay was 2 d. There were no surgery-related bleeding complications. Two cases developed surgical wound infections (methicillin-sensitive staphylococcus aureus [MSSA] and methicillin-resistant staphylococcus aureus [MRSA]) within 30 d after implantation (Table 2). One patient was explanted and eventually reimplanted and is doing quite well (patient 4, Table 2), while the other patient was explanted and has not been reimplanted. At a mean of 2 yr of follow-up, patients had approximately 40% reduction in seizure frequency, with most patients in Engel class III, indicating worthwhile seizure improvement. TABLE 2. Operative Data Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  **- Patient was non-compliant with therapy and unable to provide seizure history. MCD, malformation of cortical development; MRSA, methicillin-resistant staphylococcus aureus; MSSA, methicillin-sensitive staphylococcus aureus View Large TABLE 2. Operative Data Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  **- Patient was non-compliant with therapy and unable to provide seizure history. MCD, malformation of cortical development; MRSA, methicillin-resistant staphylococcus aureus; MSSA, methicillin-sensitive staphylococcus aureus View Large DISCUSSION Robotic-assisted surgery was first described in 1985 for a neurosurgical procedure, a thalamic biopsy, using the PUMA 560 robot.7 Despite being first used in neurosurgery, robotic platforms have been slower to become incorporated into neurosurgical procedures. While European centers have been using the ROSA robotic platform (Zimmer Biomet) in order to place SEEG electrodes for quite some time, the experience in the United States and other North American centers is relatively recent. There have been multiple studies over the last decade demonstrating robotic-assisted SEEG electrode placement to be a safe, accurate, and efficient procedure.5,8 This platform uses either markerless or scalp fiducial laser registration and can be used in either a frameless or frame-based manner. Studies have shown that registration and clinical target accuracy are best with a frame-based system, but that the ROSA robot significantly increases accuracy regardless of the system used when compared to other surface registration techniques.9,10 In our small series, we have shown robotically assisted depth electrode placement with an accuracy well within the range of what has been described in in vivo subjects in other studies.11-14 Though frame-based systems can offer similar accuracy, we have found that one main advantage to using the robot is the decrease in operating time. When we examined our operative times in robotic-assisted compared to frame-based SEEG, we found that we reduced them by 63%.5 All of the trajectory planning is done preoperatively without the need to obtain images intraoperatively or on the day of surgery. For our purposes, we frequently use trajectories orthogonal to the curvature of the skull during SEEG and will use those same trajectories for RNS placement (NeuroPace Inc). These can be particularly advantageous compared to a frame-based approach. When the patient is prone, we have found hippocampal trajectories to be quite simply placed with the robot when compared to a frame. Although there are fewer trajectories for RNS placement and thus less room for improvement, the robotic registration process is simple, occurs in the operating room with the patient in position, and adds ∼15 min to the case. This should be compared to the time necessary to place a headframe, obtain a CT (whether intra- or extraoperatively), merge the images, verify the plan, and then manually change the arc coordinates for each trajectory. Some institutions may have a protocol in place which makes the latter workflow just as timely as the former, but we have found improvements in operative time when using robotic assistance. There are very few studies detailing the operative techniques used for RNS placement1,2,10,15-17 with variable methods of reporting operative data (Table 3). Although we have not performed frame-based RNS, our nonrobotic-assisted RNS cases have typically used standard frameless stereotactic navigation. Compared to these cases, robotic-assisted operative times were slightly shorter (121 vs 151 min) while prep time was slightly longer (65 vs 41 min) so that total time from anesthesia hand off to the surgical team to the end of the procedure was similar (186 vs 192 min, data not shown). We do not have stereotactic accuracy data available for these cases but the literature suggests that frameless stereotaxy is less accurate than both frame-based and robotically assisted stereotactic cases.9,10 Thus, while total operative times and accuracy are similar for both techniques at our center, using robotic assistance should theoretically have an accuracy advantage in this example. TABLE 3. Studies Describing Surgical Techniques Used in RNS Placement Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  View Large TABLE 3. Studies Describing Surgical Techniques Used in RNS Placement Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  View Large Although it is impossible to directly compare operative techniques across centers, there are 5 other papers describing various methods of safely and accurately placing RNS devices with associated depth electrodes (Table 3). Two manuscripts did not report stereotactic accuracy or operative data.1,15 There have been 2 recent, separate descriptions of RNS placement for bilateral temporal lobe epilepsy.16,17 Both groups used skull fiducials for registration placed either pre- or intraoperatively with a subsequent CT, image fusion, and trajectory planning. Robotic assistance was used in both studies to place depth electrodes. Chan et al17 published their stereotactic accuracy and operative times with a median linear error of 2.18 mm and median operative time of 147 min in 3 patients. Rohatgi et al16 did not publish stereotactic accuracy data in their 5 patients but did publish operative times with a median of 188 min. There is only 1 study examining a frame-based operative technique for RNS without using the ROSA robot.2 The authors demonstrated a mean operative time of 163 min and achieved excellent stereotactic results with a mean Euclidian distance error of 2.67 mm. These authors also used intraoperative CT to reposition leads when necessary (2/15 leads). None of the above studies reported the time necessary to prepare the patient prior to skin incision so total operative time is impossible to compare to our study. In summary, based on the published literature, it is difficult to compare the various operative techniques and make any conclusions with regard to stereotactic accuracy, operative time, and clinical outcome. An interesting trend when reviewing the literature lies in the nature in which each center treats RNS device placement. Some studies treat RNS placement akin to deep brain stimulation with techniques aimed at maximizing stereotactic accuracy1,2 while others use techniques more often used in tumor or epilepsy surgery such as frameless neuronavigation.15 We have attempted to bridge this gap by using laser or scalp fiducial registration with robotic assistance to minimize operative time without compromising stereotactic accuracy. It should be noted, however, that the expense of robotic-assisted surgery may preclude some centers from using this technique unless and until a statistically significant advantage in either outcomes, operative time, or complications can be demonstrated. One important question to attempt to answer in the future will be the relationship between stereotactic accuracy and clinical outcome as most centers appear to highly value stereotactic accuracy. Unlike deep brain stimulation in which the relationship between clinical efficacy and electrode placement is well established,18 it is still unclear what the necessary requirements are for stereotactic accuracy and good clinical outcome in RNS. Our results (∼40% reduction in seizure frequency at 2-yr follow-up) are in line with the published trial results with errors in the 3 mm range. In addition, because infections appear to be the major complication with this technology, minimizing operative time is an important consideration. CONCLUSION The clinical application of robotic stereotactic RNS (NeuroPace Inc) electrode placement is simple and accurate. In our institution, robotic-assisted surgery is the principal operative technique for depth electrode placement in both adults and pediatric patients for SEEG and RNS.19,20 Further study in more patients with long-term follow-up is required to compare the outcomes and complications after RNS placement. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Fountas KN, Smith JR, Murro AM, Politsky J, Park YD, Jenkins PD. Implantation of a closed-loop stimulation in the management of medically refractory focal epilepsy. Stereotact Funct Neurosurg . 2005; 83( 4): 153- 158. Google Scholar CrossRef Search ADS PubMed  2. Kerolus MG, Kochanski RB, Rossi M, Stein M, Byrne RW, Sani S. Implantation of responsive neurostimulation for epilepsy using intraoperative computed tomography: technical nuances and accuracy assessment. World Neurosurg . 2017; 103: 145- 152. Google Scholar CrossRef Search ADS PubMed  3. Miller K, Halpern CH. Stereotactic bony trajectory preservation for responsive neurostimulator lead placement following depth EEG recording. Cureus . 2016; 8( 3): e549. Google Scholar PubMed  4. Diana M, Marescaux J. Robotic surgery. Br J Surg . 2015; 102( 2): e15- e28. Google Scholar CrossRef Search ADS PubMed  5. Gonzalez-Martinez J, Bulacio J, Thompson S et al.   Technique, results, and complications related to robot-assisted stereoelectroencephalography. Neurosurgery . 2016; 78( 2): 169- 180. Google Scholar CrossRef Search ADS PubMed  6. Alomar S, Jones J, Maldonado A, Gonzalez-Martinez J. The stereo-electroencephalography methodology. Neurosurg Clin North Am . 2016; 27( 1): 83- 95. Google Scholar CrossRef Search ADS   7. Kwoh YS, Hou J, Jonckheere EA, Hayati S. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng . 1988; 35( 2): 153- 160. Google Scholar CrossRef Search ADS PubMed  8. Cardinale F, Cossu M, Castana L et al.   Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery . 2013; 72( 3): 353- 366. Google Scholar CrossRef Search ADS PubMed  9. Lefranc M, Capel C, Pruvot AS et al.   The Impact of the reference imaging modality, registration method and intraoperative flat-panel computed tomography on the accuracy of the ROSA® Stereotactic Robot. Stereotact Funct Neurosurg . 2014; 92( 4): 242- 250. Google Scholar CrossRef Search ADS PubMed  10. Brandmeir NJ, Savaliya S, Rohatgi P, Sather M. The comparative accuracy of the ROSA stereotactic robot across a wide range of clinical applications and registration techniques. J Robot Surg . 2018; 12( 1): 157- 163. Google Scholar CrossRef Search ADS PubMed  11. Giese H, Hoffmann K-T, Winkelmann A, Stockhammer F, Jallo GI, Thomale U-W. Precision of navigated stereotactic probe implantation into the brainstem. J Neurosurg Pediatr . 2010; 5( 4): 350- 359. Google Scholar CrossRef Search ADS PubMed  12. Lefranc M, Capel C, Pruvot AS et al.   The impact of the reference imaging modality, registration method and intraoperative flat-panel computed tomography on the accuracy of the ROSA® stereotactic robot. Stereotact Funct Neurosurg . 2014; 92( 4): 242- 250. Google Scholar CrossRef Search ADS PubMed  13. Eljamel MS. Validation of the PathFinder™ neurosurgical robot using a phantom. Int. J. Med. Robotics Comput. Assist. Surg . 2007; 3( 4): 372- 377. Google Scholar CrossRef Search ADS   14. Pezeshkian P, DeSalles AAF, Gorgulho A, Behnke E, McArthur D, Bari A. Accuracy of frame-based stereotactic magnetic resonance imaging vs frame-based stereotactic head computed tomography fused with recent magnetic resonance imaging for postimplantation deep brain stimulator lead localization. Neurosurgery . 2011; 69( 6): 1299- 1306. Google Scholar CrossRef Search ADS PubMed  15. Lee B, Zubair MN, Marquez YD et al.   A single-center experience with the NeuroPace RNS system: a review of techniques and potential problems. World Neurosurg . 2015; 84( 3): 719- 726. Google Scholar CrossRef Search ADS PubMed  16. Rohatgi P, Jafrani RJ, Brandmeir NJ, Gilliam FG, Fisher TL, Sather MD. Technical note: robotic-guided bi-hippocampal and bi-parahippocampal depth placement for responsive neurostimulation in bitemporal lobe epilepsy. World Neurosurg . 2018; 111: 181- 189. Google Scholar CrossRef Search ADS PubMed  17. Chan AY, Mnatsakanyan L, Sazgar M et al.   Accuracy and efficacy for robotic assistance in implanting responsive neurostimulation device electrodes in bilateral mesial temporal lobe epilepsy. Operative Neurosurg . 2018; 14( 3): 267- 272. Google Scholar CrossRef Search ADS   18. Lanotte MM, Rizzone M, Bergamasco B, Faccani G, Melcarne A, Lopiano L. Deep brain stimulation of the subthalamic nucleus: anatomical, neurophysiological, and outcome correlations with the effects of stimulation. J Neurol Neurosurg Psychiatry . 2002; 72( 1): 53- 58. Google Scholar CrossRef Search ADS PubMed  19. Gonzalez-Martinez J, Bulacio J, Alexopoulos A, Jehi L, Bingaman W, Najm I. Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a North American epilepsy center. Epilepsia . 2013; 54( 2): 323- 330. Google Scholar CrossRef Search ADS PubMed  20. Gonzalez-Martinez J, Mullin J, Bulacio J et al.   Stereoelectroencephalography in children and adolescents with difficult-to-localize refractory focal epilepsy. Neurosurgery . 2014; 75( 3): 258- 268. Google Scholar CrossRef Search ADS PubMed  COMMENTS We reviewed with great interest this technical series on the placement of responsive neurostimulation leads with robotic assistance. As robotic technologies continue to advance and gain wider acceptance within neurosurgery, it is clear that they hold great potential for navigation-assistance in multi-modality epilepsy surgery for localizing temporary or permanent electrodes and laser ablation catheters at or near seizure onset zones. This retrospective series reports on an initial 12 patient experience with ROSA robot-assisted placement of responsive neurostimulator (RNS) leads with good operative efficiency (average operative time of 121 minutes) and comparable accuracy (mean Euclidean error 5.5 mm ± 1.1 mm) seen with laser surface registration techniques. We agree with the general conclusions regarding increased operative efficiency in stereotactic navigation that can be achieved with robotic assistance and welcome reports from other institutions incorporating this technology. Different robotic systems all confer unique advantages with regards to planning, set-up, registration, operative accuracy, and efficiency and will be necessary for institutions to consider data from all the major cranial robotic navigation platforms prior to adopting their own robotic systems. Comparisons to nonrobot-assisted cohorts within the same institution will also be essential in future reports to assure efficiency gains in a similar operating environment. Finally, robust and honest cost analysis of robot assisted stereotaxy compared to non-robot assisted techniques are critical for making the case for wider adoption and acceptance within neurosurgery. Allen L. Ho Gerald A. Grant Stanford, California The authors describe using robotic stereotactic navigation to place depth electrodes for responsive neurostimulatiopn. They show in their 12 patients that depth electrodes can be placed safely and efficiently as part of RNS implantation. Two of the patients developed infections requiring explantation. Larger pooled studies will be necessary to see whether robotic placement is cost effective and accurate in comparison to frame based and frameless stereotactic procedures. However, given the rapid expansion of SEEG surgery in the US, and the acquisition of surgical robots by many hospitals for this purpose, it is logical that robotic placement of RNS depth electrodes will continue to increase over time. Brett Youngerman Guy M. McKhann New York, New York Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Operative Neurosurgery Oxford University Press

Robot-Assisted Responsive Neurostimulator System Placement in Medically Intractable Epilepsy: Instrumentation and Technique

Loading next page...
 
/lp/ou_press/robot-assisted-responsive-neurostimulator-system-placement-in-3l4WRGWZpB
Publisher
Congress of Neurological Surgeons
Copyright
Copyright © 2018 by the Congress of Neurological Surgeons
ISSN
2332-4252
eISSN
2332-4260
D.O.I.
10.1093/ons/opy112
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND The management of medically refractory epilepsy patients who are not surgical candidates has remained challenging. Closed loop—or responsive—neurostimulation (RNS) is now an established therapy for the treatment of epilepsy with specific indications. The RNS® system (NeuroPace Inc, Mountainview, California) has recently been shown to be effective in reducing the seizure frequency of partial onset seizures. The electrode design consists of either intracerebral depth electrodes or subdural strip electrodes, and stereotaxis is typically used to guide placement into the EZ. Details on the operative techniques used to place these electrodes have been lacking. OBJECTIVE To address the advantage of using a robotic-assisted technique to place depth electrodes for RNS® system placement compared to the typical frame-based or frameless stereotactic systems. METHODS We retrospectively reviewed our single center, technical operative experience with RNS® system placement using robotic assistance from 2014 to 2016 via chart review. RESULTS Twelve patients underwent RNS® system placement using robotic assistance. Mean operative time was 121 min for a median of 2 depth electrodes with mean deviation from intended target of ∼3 mm in x, y, and z planes. Two patients developed wound infections, 1 of whom was reimplanted. Seizures were reduced by ∼40% at 2 yr, similar to the results seen in the open label portion of the pivotal RNS trial. CONCLUSION Robotic-assisted stereotaxis can be used to provide a stable and accurate stereotactic platform for insertion of intracerebral RNS electrodes, representing a safe, efficient and accurate procedure. Robotics, Drug-resistant epilepsy, Responsive neurostimulation, Depth electrodes, Surgical technique ABBREVIATIONS ABBREVIATIONS CT computed tomography MEG magnetoencephalography MRI magnetic resonance imaging PET positron emission tomography RNS responsive neurostimulation SEEG stereotactic electroencephalography SPECT single-photon emission computed tomography Despite the relative abundance of publications describing both the concept of closed loop stimulation and the clinical efficacy of responsive neurostimulation (RNS; NeuroPace Inc, Mountainview, California) in treating patients with partial onset focal epilepsy, there are few descriptions of the surgical technique and instrumentation necessary to properly place the depth and/or cortical strip electrodes. Surgeons typically use a frame-based system to place the leads,1,2 though frameless systems have been described.3 While popular in many other surgical subspecialties,4 robotic platforms have typically had limited applicability in most neurosurgical procedures. However, robotic-assisted placement of multiple depth electrodes for stereotactic electroencephalography (SEEG) has become increasingly utilized due to the technique's ease of use, stereotactic accuracy, and efficiency.5,6 Because many patients undergoing RNS therapy require 2 depth electrode placements, we have begun to use a robotic platform (ROSA Surgical Robot, Zimmer Biomet, Warsaw, Indiana) to place these electrodes and have found it to be just as accurate as frame-based placement and more efficient in terms of operating time. This report details our experience describing the technique of depth electrode placement using a robotic platform for RNS therapy. Table 1. Patient Characteristics Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  ATL, anterior temporal lobectomy; DTI, diffusion tensor imaging. View Large Table 1. Patient Characteristics Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  Patient  Age (yr)  Sex  Duration of epilepsy (yr)  Preop evaluation  Prior surgery  Localized epileptogenic zone  1  46  F  45  MRI, vEEG, neuropsych  B/l MT depths, L grids  Bilateral mesial temporal  2  33  F  31  MRI, vEEG, ictal SPECT, PET, neuropsych  B/l depth, L ATL  Bilateral mesial temporal  3  35  F  18  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral mesial temporal  4  42  F  7  MRI, vEEG, PET  SEEG  Bilateral hippocampus, left OF  5  21  F  5  MRI, vEEG, ictal SPECT, MEG, neuropsych  SEEG  L mesial temp (severe R hippocampal atrophy)  6  54  F  48  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG; L ATL  Bilateral mesial temporal  7  40  F  22  MRI, vEEG, ictal SPECT, MEG, PET, neuropsych  SEEG; L ATL  R mesial temporal  8  38  F  18  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral mesial temporal  9  25  M  21  MRI, vEEG, DTI  SEEG; right frontal resection  R premotor/corticospinal tract  10  47  F  16  MRI, vEEG, ictal SPECT, PET, neuropsych  Subdural strips  Bilateral hippocampal  11  26  M  21  MRI, vEEG, ictal SPECT, PET, neuropsych  SEEG  Bilateral basal temporal  12  18  M  2  MRI, vEEG, ictal SPECT, MEG  SEEG  Bilateral Heschl's gyrus  ATL, anterior temporal lobectomy; DTI, diffusion tensor imaging. View Large METHODS Patient Selection A retrospective chart review was performed in order to identify patients who had undergone robotic-assisted placement of RNS (NeuroPace Inc) therapy. This study was conducted with the approval of the Cleveland Clinic Foundation Institutional Review Board, and a waiver of informed consent was approved to allow access to protected health information by the research team. In order to be considered for RNS implantation, patients must have medically refractory simple partial or complex partial seizures, with or without secondary generalization. In order to localize the epileptic focus, our institution uses a variety of methods in the presurgical evaluation including video electroencephalography, magnetic resonance imaging (MRI), ictal single-photon emission computed tomography (SPECT), positron emission tomography (PET), and magnetoencephalography (MEG; Table 1). If the noninvasive information is discordant, invasive evaluation with SEEG or subdural grids and depth electrodes may be utilized. After this evaluation, candidates for RNS typically demonstrate bilateral mesial temporal ictal onset, unilateral mesial temporal onset with inability or unwillingness to tolerate verbal memory deficit or a focus in eloquent cortex (Table 1). Surgical Procedure The operative trajectories for depth electrode placement are planned on the robotic platform's surgical navigation software before the procedure (Figures 1A, 1C, and 2A). The patient receives prophylactic antibiotics and is placed under general anesthesia with full muscle relaxation for the entire procedure as once the patient's head position is stereotactically registered to the robot, the patient's position cannot change. For hippocampal depth electrodes, we typically place the patient in the prone position. The patient's head is fixed to the table with a standard 3-point Mayfield clamp. The robot is then registered to the head either via facial scanning using software included with the robot or using scalp fiducials if prone. Six facial landmarks are used for markerless stereotactic registration (left and right medial canthi, and lateral canthi, glabella, and nasal tip). The error is then calculated with a corrected offset value in each of the registration points. Once this error is noted to be 1 mm or less, registration is then validated and adjusted and the surgeon scans 4 facial areas (bilateral temporal and nasal areas) before the final registration verification. FIGURE 1. View largeDownload slide Robotic-assisted placement of responsive neurostimulator device. A, Illustration of surgeon planning operative trajectory on software preoperatively and hippocampal electrode placement. Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography ©2014-2017. All rights reserved. B, Preoperative MRI-planned trajectory of a hippocampal electrode used for responsive neurostimulation. C, Postoperative CT demonstrating accurate hippocampal electrode placement. FIGURE 1. View largeDownload slide Robotic-assisted placement of responsive neurostimulator device. A, Illustration of surgeon planning operative trajectory on software preoperatively and hippocampal electrode placement. Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography ©2014-2017. All rights reserved. B, Preoperative MRI-planned trajectory of a hippocampal electrode used for responsive neurostimulation. C, Postoperative CT demonstrating accurate hippocampal electrode placement. FIGURE 2. View largeDownload slide View largeDownload slide Bilateral hippocampal electrode placement for responsive neurostimulation. A, Bilateral hippocampal preoperative planning trajectories. B, Postoperative CT demonstrates accurate placement bilaterally. C, Blended CT/MRI demonstrating accurate placement bilaterally. FIGURE 2. View largeDownload slide View largeDownload slide Bilateral hippocampal electrode placement for responsive neurostimulation. A, Bilateral hippocampal preoperative planning trajectories. B, Postoperative CT demonstrates accurate placement bilaterally. C, Blended CT/MRI demonstrating accurate placement bilaterally. The surgical incisions are planned at both the electrode insertion sites and for the pulse generator. If the patient has had prior surgery, the old incision will need to be incorporated into the new incision or re-used, using the pulse generator template in order to plan an incision large enough for placement. If the patient has not had prior surgery, typically a barn door or “U” shaped incision is chosen for the pulse generator while the electrode insertion site incisions are variable depending on the electrode location. The surgical region is prepped and draped in the usual, sterile fashion. The robot is then draped under a sterile, plastic cover. A guide tube with a 2.5 mm inner cannula is attached to the robot arm. The desired trajectory is selected on the touch screen interface. After trajectory confirmation, robot arm movement is initiated through the use of a foot pedal. The robot arm locks the drilling platform into a stable position once reaching the calculated trajectory. The distance between the target site and the top of the guide tube attached to the robotic arm is pre-set at 150 mm. At the entry site, a 2.5 mm diameter hand drill bit (Stryker, Kalamazoo, Michigan) is used for making an incision and trephination in the skull. The dura is coagulated using monopolar cautery and gently punctured using a thin obturator probe. The robotic arm is then moved close to the skull to calculate a distance to the target. The electrode depth is marked and inserted. At this point, care is taken to avoid lead migration during stylet removal. A silicone cap device (NeuroPace burr hole cover model 8110; NeuroPace Inc, Mountainview, California) is placed on the bone, the plastic sheath sleeve is placed covering the electrode and secured to the bone with a titanium plate and screws. After electrode placement is completed, the robotic arm is moved out of the operative field and the process repeated for all trajectories. To place the pulse generator, the marked incision is made as outlined above. Frequently, temporalis muscle will have to be elevated in order to access the outlined region of skull. The bone is then marked out according to the dimensions of the provided ferrule, a burr hole drilled, and a craniectomy performed in the shape of the ferrule template. Care is taken to avoid penetrating through the inner table of skull, thereby potentially tearing dura or underlying cortical vessels. Bone and epidural bleeding is meticulously controlled with a combination of bonewax and hemostatic products. The provided ferrule is implanted and secured to the skull with 4 self-tapping screws (1.5-mm diameter, 4 mm long). The tunneling tool is passed such that the electrode leads are tunneled to the region of the pulse generator. The pulse generator is brought into the operative field, connected to the distal end of the depth electrodes, and secured in the ferrule. If there are additional electrodes that may be used in the future, a subgaleal pocket can be made adjacent to the burr hole site. It is important not to create a pocket near the pulse generator in order to avoid lead damage during battery replacement. The programmer wand is covered with a sterile bag to interrogate the pulse generator. The lead impedance and real-time electrocorticography are verified to ensure the leads are recording properly. The surgical wound is thoroughly irrigated and closed in anatomical layers. Dressings and a head wrap are applied. Postoperatively, a volumetric computed tomography (CT) scan (1-mm cuts) and skull X rays (anteroposterior and lateral) are obtained in the recovery room before sending the patient to the surgical floor (Figures 1B, 1D, 2B, 2C, 3B, and 3C). FIGURE 3. View largeDownload slide Placement of responsive neurostimulator electrodes in eloquent cortex. A, Preoperative MRI demonstrates prior resection just anterior to primary motor cortex. Pathology was focal cortical dysplasia and patient's seizures recurred. SEEG localized the focus to the margins of the resection cavity, and therefore responsive neurostimulator electrodes were placed around the margins of the resection, including in primary motor cortex. B and C, Postoperative skull x-ray and CT demonstrating lead placement. FIGURE 3. View largeDownload slide Placement of responsive neurostimulator electrodes in eloquent cortex. A, Preoperative MRI demonstrates prior resection just anterior to primary motor cortex. Pathology was focal cortical dysplasia and patient's seizures recurred. SEEG localized the focus to the margins of the resection cavity, and therefore responsive neurostimulator electrodes were placed around the margins of the resection, including in primary motor cortex. B and C, Postoperative skull x-ray and CT demonstrating lead placement. Data Collection and Statistical Analysis Demographic, operative, and clinical outcome data were collected from review of patients’ electronic medical records, and descriptive statistics were applied in Excel (Microsoft, Redmond, Washington). “Prep time” was considered to be the time marked as “ready for surgeon” to skin incision. “Operative time” was considered to be from skin incision until “procedure end” which is typically marked when the dressings are applied to the head. Data on the stereotactic accuracy of the procedure were collected using the robotic software (Rosana, Zimmer Biomet, Warsaw, Indiana). A postoperative CT was imported to the software and fused to the planning MRI and CT angiography. Target location errors were then calculated by measuring the distance in each plane (x, y, z) from the center of the trajectory targets and the center of the actual distal electrode contact. RESULTS In our institute from 2014 to 2016, 12 patients with localized focal epilepsy were successfully implanted with the RNS® (NeuroPace Inc) system using robotic assistance. Patient characteristics are listed in Table 1. Patients had a diagnosis of epilepsy for an average of 21 yr. Every patient had some type of prior surgery in order to attempt to localize the epileptogenic zone (either SEEG or subdural grids/strips) and 4 patients had a prior lobectomy which ultimately failed. The majority of patients had bilateral mesial temporal seizure onset zones as determined by either SEEG or depth electrodes. Table 2 demonstrates operative data and postoperative outcomes. The procedure was performed in an average operative time of 121 min (range 70-184 min) with 65 min of prep time (range 42-98 min) for a median of 2 electrodes per surgery. Twenty-six total electrodes were placed. The mean absolute error for electrode placement was 3.2 ± 0.8 mm in the x plane, 3.4 ± 0.6 mm in the y plane, and 2.9 ± 0.6 mm in the z plane such that the mean Euclidean error was 5.5 ± 1.1 mm. All patients did well immediately postoperatively without any neurological decline or other complications. The median hospital stay was 2 d. There were no surgery-related bleeding complications. Two cases developed surgical wound infections (methicillin-sensitive staphylococcus aureus [MSSA] and methicillin-resistant staphylococcus aureus [MRSA]) within 30 d after implantation (Table 2). One patient was explanted and eventually reimplanted and is doing quite well (patient 4, Table 2), while the other patient was explanted and has not been reimplanted. At a mean of 2 yr of follow-up, patients had approximately 40% reduction in seizure frequency, with most patients in Engel class III, indicating worthwhile seizure improvement. TABLE 2. Operative Data Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  **- Patient was non-compliant with therapy and unable to provide seizure history. MCD, malformation of cortical development; MRSA, methicillin-resistant staphylococcus aureus; MSSA, methicillin-sensitive staphylococcus aureus View Large TABLE 2. Operative Data Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  Patient  Targets  Electrode #  Absolute mean error – x (mm)  Absolute mean error – y (mm)  Absolute mean error – z (mm)  Mean Euclidian error (mm)  Prep time (min)  Operative time (min)  Length of stay (d)  Seizure outcome (Engel)  % Seizure reduction  Follow-up (mo)  Complications  1  1. Amygdala 2. Hippocampus  2  2.7  2.7  4.2  5.7  49  118  2  IB  100  36. 7  None  2  1. Amygdala 2. Hippocampus 3. Parahippocampal gyrus  3  0.6  0.4  0.6  0.9  98  177  2  IIIA  72  35.7  None  3  1. Left hippocampus 2. Right hippocampus  2  1.8  5.8  4.6  7.6  72  127  2  IVA  10  34  None  4  1. Left hippocampus 2. Right hippocampus 3. Left orbitofrontal cortex  3  11.7  8.3  8.1  16.5  45  184  2  IIB  94  29.7  MRSA wound infection  5  1. Amygdala 2. Hippocampus  2  2.6  0.7  2.1  3.4  78  106  2  IIIA  75  33.7  None  6  1. Amygdala 2. Hippocampus  2  2.4  11  6.1  12.8  67  70  2  IVA  25  34.5  None  7  1. Hippocampus 2. Parahippocampal gyrus  2  0.4  5.3  2.6  5.9  76  105  2  **  **  24.3  None  8  1. Left hippocampus 2. Right hippocampus  2  3.6  1.5  3.1  5.0  62  119  3  IVB  0  12.4  MSSA wound infection  9  1. Anterior to MCD lesion 2. Posterior to MCD lesion 3. Lateral to MCD lesion  3  2.6  2.8  2.2  4.4  42  132  2  IIIA  50  24.2  Numbness/tingling L 3rd, 4th, 5th digits  10  1. Left hippocampus 2. Right hippocampus  2  3.2  0.7  1.2  3.5  69  104  3  IVB  0  9.8  None  11  1. Left basal temporal 2. Right basal temporal  2  2.1  2.2  0.6  3.1  64  103  1  IVC  –33  9.5  None  12  1. Left Heschl's gyrus 2. Right Heschl's gyrus  2  1.5  2  0.4  2.5  58  111  2  IIIA  50  7  None  **- Patient was non-compliant with therapy and unable to provide seizure history. MCD, malformation of cortical development; MRSA, methicillin-resistant staphylococcus aureus; MSSA, methicillin-sensitive staphylococcus aureus View Large DISCUSSION Robotic-assisted surgery was first described in 1985 for a neurosurgical procedure, a thalamic biopsy, using the PUMA 560 robot.7 Despite being first used in neurosurgery, robotic platforms have been slower to become incorporated into neurosurgical procedures. While European centers have been using the ROSA robotic platform (Zimmer Biomet) in order to place SEEG electrodes for quite some time, the experience in the United States and other North American centers is relatively recent. There have been multiple studies over the last decade demonstrating robotic-assisted SEEG electrode placement to be a safe, accurate, and efficient procedure.5,8 This platform uses either markerless or scalp fiducial laser registration and can be used in either a frameless or frame-based manner. Studies have shown that registration and clinical target accuracy are best with a frame-based system, but that the ROSA robot significantly increases accuracy regardless of the system used when compared to other surface registration techniques.9,10 In our small series, we have shown robotically assisted depth electrode placement with an accuracy well within the range of what has been described in in vivo subjects in other studies.11-14 Though frame-based systems can offer similar accuracy, we have found that one main advantage to using the robot is the decrease in operating time. When we examined our operative times in robotic-assisted compared to frame-based SEEG, we found that we reduced them by 63%.5 All of the trajectory planning is done preoperatively without the need to obtain images intraoperatively or on the day of surgery. For our purposes, we frequently use trajectories orthogonal to the curvature of the skull during SEEG and will use those same trajectories for RNS placement (NeuroPace Inc). These can be particularly advantageous compared to a frame-based approach. When the patient is prone, we have found hippocampal trajectories to be quite simply placed with the robot when compared to a frame. Although there are fewer trajectories for RNS placement and thus less room for improvement, the robotic registration process is simple, occurs in the operating room with the patient in position, and adds ∼15 min to the case. This should be compared to the time necessary to place a headframe, obtain a CT (whether intra- or extraoperatively), merge the images, verify the plan, and then manually change the arc coordinates for each trajectory. Some institutions may have a protocol in place which makes the latter workflow just as timely as the former, but we have found improvements in operative time when using robotic assistance. There are very few studies detailing the operative techniques used for RNS placement1,2,10,15-17 with variable methods of reporting operative data (Table 3). Although we have not performed frame-based RNS, our nonrobotic-assisted RNS cases have typically used standard frameless stereotactic navigation. Compared to these cases, robotic-assisted operative times were slightly shorter (121 vs 151 min) while prep time was slightly longer (65 vs 41 min) so that total time from anesthesia hand off to the surgical team to the end of the procedure was similar (186 vs 192 min, data not shown). We do not have stereotactic accuracy data available for these cases but the literature suggests that frameless stereotaxy is less accurate than both frame-based and robotically assisted stereotactic cases.9,10 Thus, while total operative times and accuracy are similar for both techniques at our center, using robotic assistance should theoretically have an accuracy advantage in this example. TABLE 3. Studies Describing Surgical Techniques Used in RNS Placement Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  View Large TABLE 3. Studies Describing Surgical Techniques Used in RNS Placement Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  Authors/Center  No. of patients  Stereotactic method  Registration method  No of electrodes  Accuracy, mean linear or Euclidian distance, mm, (median)  Mean operative time, minutes, (median)  Outcome  Mean follow-up time, months (median)  Fountas et al1/MCG  8  Leksell frame based  CT with affixed frame  15  n/a  n/a  7/8 with >45% seizure reduction  9.3 (7)  Lee et al15/USC  10  Frameless neuronavigation  n/a  20  n/a  n/a  11% seizure frequency reduction  12  Chan et al17/UCI  3  Robotic-assisted  Skull fiducials with intraoperative CT  6  2.28 (2.18)  149 (147)  80% seizure frequency reduction (median)  3-6  Rohatgi et al16/Penn State  5  Robotic-assisted  Skull fiducials with preoperative CT  10  n/a  224 (188)  40% Engel class I, 40% Engel class III, 20% Engel class IV  21.2 (20)  Kerolus et al2/Rush  8  Leksell frame-based  CT with affixed frame, intraoperative CT used for trajectory changes  15  2.63 (2.53)  163 (168)  74% seizure frequency reduction  9 (7.8)  McGovern et al/CCF (present study)  12  Robotic-assisted  Laser facial recognition or scalp fiducials  26  3.16 (5.5)  121 (115)  40% seizure frequency reduction  24 (26.8)  View Large Although it is impossible to directly compare operative techniques across centers, there are 5 other papers describing various methods of safely and accurately placing RNS devices with associated depth electrodes (Table 3). Two manuscripts did not report stereotactic accuracy or operative data.1,15 There have been 2 recent, separate descriptions of RNS placement for bilateral temporal lobe epilepsy.16,17 Both groups used skull fiducials for registration placed either pre- or intraoperatively with a subsequent CT, image fusion, and trajectory planning. Robotic assistance was used in both studies to place depth electrodes. Chan et al17 published their stereotactic accuracy and operative times with a median linear error of 2.18 mm and median operative time of 147 min in 3 patients. Rohatgi et al16 did not publish stereotactic accuracy data in their 5 patients but did publish operative times with a median of 188 min. There is only 1 study examining a frame-based operative technique for RNS without using the ROSA robot.2 The authors demonstrated a mean operative time of 163 min and achieved excellent stereotactic results with a mean Euclidian distance error of 2.67 mm. These authors also used intraoperative CT to reposition leads when necessary (2/15 leads). None of the above studies reported the time necessary to prepare the patient prior to skin incision so total operative time is impossible to compare to our study. In summary, based on the published literature, it is difficult to compare the various operative techniques and make any conclusions with regard to stereotactic accuracy, operative time, and clinical outcome. An interesting trend when reviewing the literature lies in the nature in which each center treats RNS device placement. Some studies treat RNS placement akin to deep brain stimulation with techniques aimed at maximizing stereotactic accuracy1,2 while others use techniques more often used in tumor or epilepsy surgery such as frameless neuronavigation.15 We have attempted to bridge this gap by using laser or scalp fiducial registration with robotic assistance to minimize operative time without compromising stereotactic accuracy. It should be noted, however, that the expense of robotic-assisted surgery may preclude some centers from using this technique unless and until a statistically significant advantage in either outcomes, operative time, or complications can be demonstrated. One important question to attempt to answer in the future will be the relationship between stereotactic accuracy and clinical outcome as most centers appear to highly value stereotactic accuracy. Unlike deep brain stimulation in which the relationship between clinical efficacy and electrode placement is well established,18 it is still unclear what the necessary requirements are for stereotactic accuracy and good clinical outcome in RNS. Our results (∼40% reduction in seizure frequency at 2-yr follow-up) are in line with the published trial results with errors in the 3 mm range. In addition, because infections appear to be the major complication with this technology, minimizing operative time is an important consideration. CONCLUSION The clinical application of robotic stereotactic RNS (NeuroPace Inc) electrode placement is simple and accurate. In our institution, robotic-assisted surgery is the principal operative technique for depth electrode placement in both adults and pediatric patients for SEEG and RNS.19,20 Further study in more patients with long-term follow-up is required to compare the outcomes and complications after RNS placement. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Fountas KN, Smith JR, Murro AM, Politsky J, Park YD, Jenkins PD. Implantation of a closed-loop stimulation in the management of medically refractory focal epilepsy. Stereotact Funct Neurosurg . 2005; 83( 4): 153- 158. Google Scholar CrossRef Search ADS PubMed  2. Kerolus MG, Kochanski RB, Rossi M, Stein M, Byrne RW, Sani S. Implantation of responsive neurostimulation for epilepsy using intraoperative computed tomography: technical nuances and accuracy assessment. World Neurosurg . 2017; 103: 145- 152. Google Scholar CrossRef Search ADS PubMed  3. Miller K, Halpern CH. Stereotactic bony trajectory preservation for responsive neurostimulator lead placement following depth EEG recording. Cureus . 2016; 8( 3): e549. Google Scholar PubMed  4. Diana M, Marescaux J. Robotic surgery. Br J Surg . 2015; 102( 2): e15- e28. Google Scholar CrossRef Search ADS PubMed  5. Gonzalez-Martinez J, Bulacio J, Thompson S et al.   Technique, results, and complications related to robot-assisted stereoelectroencephalography. Neurosurgery . 2016; 78( 2): 169- 180. Google Scholar CrossRef Search ADS PubMed  6. Alomar S, Jones J, Maldonado A, Gonzalez-Martinez J. The stereo-electroencephalography methodology. Neurosurg Clin North Am . 2016; 27( 1): 83- 95. Google Scholar CrossRef Search ADS   7. Kwoh YS, Hou J, Jonckheere EA, Hayati S. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng . 1988; 35( 2): 153- 160. Google Scholar CrossRef Search ADS PubMed  8. Cardinale F, Cossu M, Castana L et al.   Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery . 2013; 72( 3): 353- 366. Google Scholar CrossRef Search ADS PubMed  9. Lefranc M, Capel C, Pruvot AS et al.   The Impact of the reference imaging modality, registration method and intraoperative flat-panel computed tomography on the accuracy of the ROSA® Stereotactic Robot. Stereotact Funct Neurosurg . 2014; 92( 4): 242- 250. Google Scholar CrossRef Search ADS PubMed  10. Brandmeir NJ, Savaliya S, Rohatgi P, Sather M. The comparative accuracy of the ROSA stereotactic robot across a wide range of clinical applications and registration techniques. J Robot Surg . 2018; 12( 1): 157- 163. Google Scholar CrossRef Search ADS PubMed  11. Giese H, Hoffmann K-T, Winkelmann A, Stockhammer F, Jallo GI, Thomale U-W. Precision of navigated stereotactic probe implantation into the brainstem. J Neurosurg Pediatr . 2010; 5( 4): 350- 359. Google Scholar CrossRef Search ADS PubMed  12. Lefranc M, Capel C, Pruvot AS et al.   The impact of the reference imaging modality, registration method and intraoperative flat-panel computed tomography on the accuracy of the ROSA® stereotactic robot. Stereotact Funct Neurosurg . 2014; 92( 4): 242- 250. Google Scholar CrossRef Search ADS PubMed  13. Eljamel MS. Validation of the PathFinder™ neurosurgical robot using a phantom. Int. J. Med. Robotics Comput. Assist. Surg . 2007; 3( 4): 372- 377. Google Scholar CrossRef Search ADS   14. Pezeshkian P, DeSalles AAF, Gorgulho A, Behnke E, McArthur D, Bari A. Accuracy of frame-based stereotactic magnetic resonance imaging vs frame-based stereotactic head computed tomography fused with recent magnetic resonance imaging for postimplantation deep brain stimulator lead localization. Neurosurgery . 2011; 69( 6): 1299- 1306. Google Scholar CrossRef Search ADS PubMed  15. Lee B, Zubair MN, Marquez YD et al.   A single-center experience with the NeuroPace RNS system: a review of techniques and potential problems. World Neurosurg . 2015; 84( 3): 719- 726. Google Scholar CrossRef Search ADS PubMed  16. Rohatgi P, Jafrani RJ, Brandmeir NJ, Gilliam FG, Fisher TL, Sather MD. Technical note: robotic-guided bi-hippocampal and bi-parahippocampal depth placement for responsive neurostimulation in bitemporal lobe epilepsy. World Neurosurg . 2018; 111: 181- 189. Google Scholar CrossRef Search ADS PubMed  17. Chan AY, Mnatsakanyan L, Sazgar M et al.   Accuracy and efficacy for robotic assistance in implanting responsive neurostimulation device electrodes in bilateral mesial temporal lobe epilepsy. Operative Neurosurg . 2018; 14( 3): 267- 272. Google Scholar CrossRef Search ADS   18. Lanotte MM, Rizzone M, Bergamasco B, Faccani G, Melcarne A, Lopiano L. Deep brain stimulation of the subthalamic nucleus: anatomical, neurophysiological, and outcome correlations with the effects of stimulation. J Neurol Neurosurg Psychiatry . 2002; 72( 1): 53- 58. Google Scholar CrossRef Search ADS PubMed  19. Gonzalez-Martinez J, Bulacio J, Alexopoulos A, Jehi L, Bingaman W, Najm I. Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a North American epilepsy center. Epilepsia . 2013; 54( 2): 323- 330. Google Scholar CrossRef Search ADS PubMed  20. Gonzalez-Martinez J, Mullin J, Bulacio J et al.   Stereoelectroencephalography in children and adolescents with difficult-to-localize refractory focal epilepsy. Neurosurgery . 2014; 75( 3): 258- 268. Google Scholar CrossRef Search ADS PubMed  COMMENTS We reviewed with great interest this technical series on the placement of responsive neurostimulation leads with robotic assistance. As robotic technologies continue to advance and gain wider acceptance within neurosurgery, it is clear that they hold great potential for navigation-assistance in multi-modality epilepsy surgery for localizing temporary or permanent electrodes and laser ablation catheters at or near seizure onset zones. This retrospective series reports on an initial 12 patient experience with ROSA robot-assisted placement of responsive neurostimulator (RNS) leads with good operative efficiency (average operative time of 121 minutes) and comparable accuracy (mean Euclidean error 5.5 mm ± 1.1 mm) seen with laser surface registration techniques. We agree with the general conclusions regarding increased operative efficiency in stereotactic navigation that can be achieved with robotic assistance and welcome reports from other institutions incorporating this technology. Different robotic systems all confer unique advantages with regards to planning, set-up, registration, operative accuracy, and efficiency and will be necessary for institutions to consider data from all the major cranial robotic navigation platforms prior to adopting their own robotic systems. Comparisons to nonrobot-assisted cohorts within the same institution will also be essential in future reports to assure efficiency gains in a similar operating environment. Finally, robust and honest cost analysis of robot assisted stereotaxy compared to non-robot assisted techniques are critical for making the case for wider adoption and acceptance within neurosurgery. Allen L. Ho Gerald A. Grant Stanford, California The authors describe using robotic stereotactic navigation to place depth electrodes for responsive neurostimulatiopn. They show in their 12 patients that depth electrodes can be placed safely and efficiently as part of RNS implantation. Two of the patients developed infections requiring explantation. Larger pooled studies will be necessary to see whether robotic placement is cost effective and accurate in comparison to frame based and frameless stereotactic procedures. However, given the rapid expansion of SEEG surgery in the US, and the acquisition of surgical robots by many hospitals for this purpose, it is logical that robotic placement of RNS depth electrodes will continue to increase over time. Brett Youngerman Guy M. McKhann New York, New York Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Operative NeurosurgeryOxford University Press

Published: May 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off