Development of a novel frameless skull-mounted ball-joint guide array for use in image-guided neurosurgery

Vivek Sudhakar Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Vivek Sudhakar in
Current site
Google Scholar
PubMed
Close
 BS
,
Amin Mahmoodi Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Amin Mahmoodi in
Current site
Google Scholar
PubMed
Close
 BS
,
John R. Bringas Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by John R. Bringas in
Current site
Google Scholar
PubMed
Close
 BS
,
Jerusha Naidoo Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Jerusha Naidoo in
Current site
Google Scholar
PubMed
Close
 PhD
,
Adrian Kells Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Adrian Kells in
Current site
Google Scholar
PubMed
Close
 PhD
,
Lluis Samaranch Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Lluis Samaranch in
Current site
Google Scholar
PubMed
Close
 PhD
,
Massimo S. Fiandaca Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Massimo S. Fiandaca in
Current site
Google Scholar
PubMed
Close
 MD
, and
Krystof S. Bankiewicz Interventional Neuro Center, Department of Neurological Surgery, University of California San Francisco, California

Search for other papers by Krystof S. Bankiewicz in
Current site
Google Scholar
PubMed
Close
 MD, PhD
Full access

OBJECTIVE

Successful convection-enhanced delivery of therapeutic agents to subcortical brain structures requires accurate cannula placement. Stereotactic guiding devices have been developed to accurately target brain nuclei. However, technologies remain limited by a lack of MRI compatibility, or by devices’ size, making them suboptimal for direct gene delivery to brain parenchyma. The goal of this study was to validate the accuracy of a novel frameless skull-mounted ball-joint guide array (BJGA) in targeting the nonhuman primate (NHP) brain.

METHODS

Fifteen MRI-guided cannula insertions were performed on 9 NHPs, each targeting the putamen. Optimal trajectories were planned on a standard MRI console using 3D multiplanar baseline images. After cannula insertion, the intended trajectory was compared to the final trajectory to assess deviation (euclidean error) of the cannula tip.

RESULTS

The average cannula tip deviation was 1.18 ± 0.60 mm (mean ± SD) as measured by 2 independent reviewers. Topological analysis showed a superior, posterior, and rightward directional bias, and the intra- and interclass correlation coefficients were > 0.85, indicating valid and reliable intra- and interobserver evaluation.

CONCLUSIONS

The data demonstrate that the BJGA can be used to reliably target subcortical brain structures by using MRI guidance, with accuracy comparable to current frameless stereotactic systems. The size and versatility of the BJGA, combined with a streamlined workflow, allows for its potential applicability to a variety of intracranial neurosurgical procedures, and for greater flexibility in executing MRI-guided experiments within the NHP brain.

ABBREVIATIONS

AC-PC = anterior commissure–posterior commissure; BJGA = ball-joint guide array; CED = convection-enhanced delivery; DICOM = Digital Imaging and Communications in Medicine; iMRI = intraoperative MRI; MPR = multiplanar reconstruction; NHP = nonhuman primate; PEEK = polyetheretherketone; T1W = T1-weighted.

OBJECTIVE

Successful convection-enhanced delivery of therapeutic agents to subcortical brain structures requires accurate cannula placement. Stereotactic guiding devices have been developed to accurately target brain nuclei. However, technologies remain limited by a lack of MRI compatibility, or by devices’ size, making them suboptimal for direct gene delivery to brain parenchyma. The goal of this study was to validate the accuracy of a novel frameless skull-mounted ball-joint guide array (BJGA) in targeting the nonhuman primate (NHP) brain.

METHODS

Fifteen MRI-guided cannula insertions were performed on 9 NHPs, each targeting the putamen. Optimal trajectories were planned on a standard MRI console using 3D multiplanar baseline images. After cannula insertion, the intended trajectory was compared to the final trajectory to assess deviation (euclidean error) of the cannula tip.

RESULTS

The average cannula tip deviation was 1.18 ± 0.60 mm (mean ± SD) as measured by 2 independent reviewers. Topological analysis showed a superior, posterior, and rightward directional bias, and the intra- and interclass correlation coefficients were > 0.85, indicating valid and reliable intra- and interobserver evaluation.

CONCLUSIONS

The data demonstrate that the BJGA can be used to reliably target subcortical brain structures by using MRI guidance, with accuracy comparable to current frameless stereotactic systems. The size and versatility of the BJGA, combined with a streamlined workflow, allows for its potential applicability to a variety of intracranial neurosurgical procedures, and for greater flexibility in executing MRI-guided experiments within the NHP brain.

In Brief

A novel guiding device has been developed that allows for accurate targeting of specific brain regions. The small profile and unique design of the device allow for accurate targeting of specific brain regions without the requirement of specialized MR software.

Efficient and accurate delivery of therapeutic agents to specific brain regions is central to the success of many gene and cell-based therapies currently in preclinical and clinical development. Open microsurgical1,21 and stereotactic2 surgical interventions allow precise anatomical targeting, and parenchymal delivery of molecular therapies bypasses the chemical and structural constraints imposed by the blood-brain barrier.29,30 Recent improvements in stereotactic surgical procedures and an increased understanding of the pathophysiology of neurological disorders are expected to make direct intracranial delivery of therapeutic agents a viable treatment option for many neurological diseases.

The past decade has seen a shift toward frameless guidance technologies for neurosurgical biopsies, microelectrode recordings, and cannula placements.17,24 Frameless platforms are lighter, smaller, and allow increased access to the surgical field. Intraoperative assessment of lightly sedated or awake patients is easier, and patients are less apprehensive about being fitted with a small intraoperative skull-mounted guidance device than with larger traditional stereotactic frames.14,19 Concurrently there has been an increase in the use of intraoperative MRI (iMRI), increasing the need for MRI-compatible navigational devices. Because most older frameless systems are not MRI compatible,23 two frameless, MRI-compatible platforms have been developed for preclinical and clinical gene therapy procedures: the Nexframe system (Medtronic, Inc.) and the ClearPoint system (MRI Interventions, Inc.). These devices are useful for targeting certain brain structures; however, they require specialized MRI software and their large profile imposes limitations on efficient targeting of deeper brain structures.

Our experience with real-time convection-enhanced delivery (CED) protocols in nonhuman primates (NHPs) and humans4,9–12,32–35,38,41 has led us to develop a novel frameless guidance system: the skull-mounted ball-joint guide array (BJGA). The BJGA is smaller, features a wider range of motion, and is compatible with most MRI platforms and software. This study describes the device in detail and documents the targeting accuracy of the BJGA within the NHP brain.

Methods

Animals

Nine adult male rhesus macaques (Macaca mulatta, 3–7 kg) were included in this study. All procedures were carried out with the approval of the Institutional Animal Care and Use Committee (IACUC) and in accordance with the standard operating procedures at our institution. All animal care complied with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council).

Guidance System

Ball-Joint Guide Array

The implantable BJGA (Fig. 1A–E) consists of the following: a threaded base fitted with 3 holes for securing the device to the skull; a 3-hole ball array with locking side screw (the internal diameter of each circular guidance port is 2 mm and accepts up to a 16-gauge cannula); a knurled threaded collar that releases or locks the array to the threaded base; and a 3-prong contrast-filled fiducial. Each fiducial contains 3 parallel tubes filled with a 2-mM gadolinium-based solution that appears hyperintense on T1-weighted (T1W) images (Fig. 1E). The BJGA was fabricated using polyetheretherketone (PEEK), and was designed to accept infusion cannulae, electrodes, and biopsy needles. The BJGA weighs 4.30 g, has a maximum angulation of 16° from the vertical, with a full 360° rotation over the threaded base (Fig. 1F–H), and can be used with any onboard MRI software.

FIG. 1.
FIG. 1.

Schematics of the BJGA: measurements, implantation landmarks, and device assembly. A: Main parts and dimensions of the BJGA. The 3-hole ball array with adjustable side screw (black dot) is the core of the device, and guides the cannula to the targeted structure (1). The threaded skull-mounted base is secured onto the skull by 3 titanium fasteners (2). The knurled threaded locking collar interlocks with the base and secures the device in the desired position (3). B: Apical view of the ball array and the skull-mounted base with the most relevant measurements. C: Assembly of the BJGA with all parts. D: Another essential piece of the BJGA is the gadolinium-based fluid–filled guiding tubing that is placed inside the ball array during trajectory planning (1). As depicted in the cross-section diagram (2), this fiducial consists of a threaded cap (red arrowheads), a small chamber (blue arrowheads), and 3 PEEK tubes (green arrowheads) filled with a 2-mM solution of a gadolinium-based MR contrast agent. E: Representative T1W MR image of the BJGA and the fiducial tubes, demonstrating the gadolinium signal from the chamber and the tubing. The BJGA is not visible on MRI and has been contoured with dashed lines. Red arrowheads represent the threaded cap, blue arrowheads represent the gadolinium-filled chamber, and green arrowheads represent the PEEK tubes. F: Lateral view of the BJGA on an NHP skull model with the fiducials inserted in the ball. The maximum angulation of 16° is shown in red. G: Superior view of bilateral implantation of the BJGA on an NHP skull model with the fiducials inserted in the ball. The maximum angulation of 16° is shown in red. H: Posterior view of the BJGA with (right) and without (left) the locking collar. The maximum angulation of 16° is shown in red. I: Posterior view of an NHP skull with the skull-mounted base placed bilaterally. Relevant anatomical landmarks on the NHP skull model—the midline (black arrowheads) and the nuchal ridge (white arrowheads), are noted.

Image-Guided Preplanning

Optimal burr hole placement and a preliminary trajectory were generated prior to surgery with Horos, an open-source Digital Imaging and Communications in Medicine (DICOM) viewer. The position of each burr hole was optimized for accessing the long axis of the putamen (Fig. 1I). Medial-lateral and superior-inferior measurements were made on baseline MR images and the distance from the intersection of the midline and nuchal ridge was recorded. Once the burr hole position was determined, a preliminary trajectory was drawn.

Implantation Surgery and Catheter Placement

Implantation Surgery

Animals were sedated with an intramuscular injection of ketamine (10 mg/kg) and medetomidine (0.015 mg/kg). After orotracheal intubation, the animals were provided with inhalational isoflurane (1%–4%) and closely monitored by a veterinary anesthetist. Animals were positioned prone, with their heads centered, slightly flexed, and secured within an MRI-compatible headholder for easier access to the occiput.

A transverse occipital scalp incision was made just caudal to the bony nuchal ridge. The incision location was selected to prevent undue pressure by the nuchal ridge on the incision, potentially complicating postoperative wound healing. The scalp and galea were mobilized superiorly, and the pericranium was reflected to expose the occipital skull. Caudal to the nuchal ridge, the cervical muscle skull attachments were released to expose the underlying skull to a predetermined distance from the midline, allowing optimal burr hole placement (above or on the nuchal ridge and lateral to the midline; see Fig. 1I). Two small burr holes were made over each occipital region of the skull, leaving the underlying dura mater intact. Astride each burr hole, the threaded base of the guidance system was attached to the skull with 3 MRI-compatible titanium self-drilling microplate screws (Fig. 1I). The BJGA was mounted onto the base and locked in place using the knurled locking collar. The wound site was closed in anatomical layers around the BJGA and a sterile adhesive (Ioban; 3M, Inc.) was positioned over the skin to maintain sterility. The BJGA-implanted animal was then transferred to the MRI scanner, where baseline images were acquired for trajectory planning.

MRI Acquisition

For our study, all scanning procedures were performed on a Siemens 3-T Magnetom Prisma (Siemens AG) by using 2 Siemens 3-T loop coils (11 cm) placed in parallel and secured on the lateral aspect of each animal’s head. Three-dimensional, T1W, magnetization-prepared rapid gradient echo (3D-MPRAGE) images were obtained with TR 2110 msec, TE 3.6 msec, flip angle 15°, number of excitations 1, matrix 240 × 240, field of view 240 × 240 × 240, and slice thickness 1 mm. These parameters resulted in a 1-mm3 voxel volume.

Trajectory Planning

The animal was positioned on the MRI scanning bed. Contrast-containing fiducials were placed into each BJGA, and the animal was advanced into the MRI gantry for baseline midbore image acquisition. Volumetric T1W images were acquired with a field of view covering the complete NHP brain and the skull-mounted BJGAs. Cannula trajectories were planned on the Siemens 3-T Magnetom Prisma console with the oblique multiplanar reconstruction (3D-MPR) function of the Siemens MRI software. Before trajectories can be planned, it is necessary to achieve complete visualization of the fiducials in both the oblique axial and sagittal planes, identifying their trajectory relative to the proposed subcortical targets (Fig. 2). Window settings were adjusted to optimally visualize the gadolinium signal in each fiducial. Once this was achieved, medial-lateral, anterior-posterior, and rotational adjustments were made to the BJGA stem to align it with the proposed target site (Fig. 3A–C). T1W images acquired between each adjustment verified optimal alignment of the BJGA for target acquisition. The proper access hole was then determined (Fig. 3D) and a trajectory was drawn. A safety check was done to ensure that the planned trajectory avoided critical structures (e.g., vasculature in sulci, off-target parenchyma, and cerebral ventricles) (Fig. 3E and F). Once the trajectory was optimized, the BJGA was locked in position. Measurements were made on the Prisma console to determine the insertion depth of each cannula, using the body of the fiducial as a surrogate for the top of the BJGA stem (Fig. 3E and F, white arrowhead). We then subtracted 3 mm from the total distance to account for the distance between the end of the gadolinium cartridge and the top of the ball array (see Fig. 1E).

FIG. 2.
FIG. 2.

Workflow for the digital alignment of the fiducial’s gadolinium signal of the guiding device using the MPR toolbox. Before target selection and trajectory planning, the xyz planes must be oriented such that the entire fiducial length is visible in the MRI sequences. First, T1W MRI sequences were acquired, and a portion of the gadolinium-positive fiducial tubes was visualized (white arrowhead). Step 1: The x plane is oriented on top of the midline (axial); the z plane is oriented parallel to the AC-PC line (sagittal). Step 2: The crosshair (white circle) is placed on top of one of the fiducial signals (straight arrow), and the z plane is oriented to the fiducial signal (sagittal; curved arrow). Step 3: The x plane is rotated to align with the gadolinium signal in the axial plane (curved arrow). Small refinements can be done in the sagittal plane to correctly align the z plane to the fiducial signal. Step 4: The crosshair is located in the center of the fiducial tube in the coronal view. Numbers 1–3 label each of the 3 holes of the ball array.

FIG. 3.
FIG. 3.

Schematic workflow for appropriate BJGA manipulation to the selected target (i.e., putamen). Once the BJGA is mounted and secured on the skull, the array can be manually adjusted to target the selected structure. Initially, if the array is not targeting the structure (A, white arrowheads depict missing putamen) multiple adjustments in the medial-lateral and superior-inferior direction can be made until the trajectory targets the structure (B [left panel, white arrowhead], shows the trajectory medial to the putamen, and B [right panel, white arrowhead], shows the trajectory appropriately targeting the putamen). Once the trajectory is defined (C), the surgeon selects the appropriate access hole (D). The surgeon must then ensure that the trajectory is not proximal to sulci or ventricles, to avoid a potential large-vessel insult. If the trajectory is not satisfactory, the surgeon can manually adjust the BJGA to redefine the trajectory. Image (E) shows an example of suboptimal trajectory that crosses too close to a sulcus (white arrowhead). Image (F) shows how the previous trajectory is altered by superiorly adjusting the BJGA, and changing the entry hole from lateral to medial (insets, panels E and F). The new trajectory satisfies the safety check with regard to the sulcus (white arrowhead). The surgeon then measures the depth from the bottom of the gadolinium-filled chamber (white arrow) to the target point (i.e., putamen), subtracts 3 mm to account for the distance between the bottom of the chamber and the top of the BJGA (see Fig. 1D and 1E), and marks the distance on the cannula. A depth stop is placed at the target mark, the cannula is introduced through the selected entry hole to the target, and the cannula location is determined (G). If the cannula tip is in an optimal position, the infusion can begin (H).

Cannula Placement

The animal was removed from the interior of the MRI gantry to facilitate surgical access to the BJGA. The surgeon then marked the distance from the proximal tip of a 16-gauge, fused silica, reflux-resistant ceramic cannula32 (SmartFlow catheter; MRI Interventions), and a depth stop was attached to prevent unwanted advancement of the cannula. First, a MRI-compatible lancet was introduced through the selected guidance port on the BJGA stem to pierce the dura. The lancet was then removed, and the premeasured ceramic cannula was then slowly inserted through the BJGA to the proposed target depth. Attention was paid to avoid rotation of the cannula during placement—insertional rotation or lateral deviation may compromise the parenchymal seal provided and increase the likelihood of infusate reflux. Cannula tubing for infusions was secured to the headholder, and the animal was advanced into the MRI gantry for T1W imaging (Fig. 3G and H).

MRI Data

Imaging Data Analysis

The DICOM data were viewed using an open-source DICOM reader and imaging software, Horos (https://horosproject.org/). Horos was used to calculate the discrepancy between the planned and observed cannula tip location. A 3D-MPR data set was created and manipulated, such that an oblique projection of the baseline MRI DICOM data matched the planned trajectory. The cannula tip was then identified on the baseline 3D-MPR based on the planned depth of infusion. An x, y, z coordinate of the planned cannula tip location was marked on the axial, sagittal, and coronal slices of the baseline 3D-MPR. Postinsertion cannula positioning was determined from 3D-MPR scans that were obliquely reconstructed to clearly depict the entire cannula path. The actual ceramic cannula tip location was marked in the axial, coronal, and sagittal planes, and its x, y, z coordinates were determined. The anterior commissure–posterior commissure (AC-PC) line was delineated as previously described11 and the midcommissural point identified. The Horos-defined coordinates were then mapped into the AC-PC plane, defined as a coordinate grid in which the midpoint of the AC-PC line—the midcommissural point—is set as the origin with coordinates of 0, 0, 0 (Equation 1).

In this equation, x, y, and z are the cannula tip coordinates and dx, dy, and dz are the coordinates of the midcommissural point. The resulting matrix product defines the target (x′, y′, z′) coordinates in AC-PC space.

The newly mapped points were then used to determine the euclidean error, which was calculated using the standard euclidean equation for distances in 3D space (Equation 2).

In this equation: (x′2 − x′1) is defined as the medial-lateral, (y′2 − y′1) as the anterior-posterior, and (z′2 − z′1) as the superior-inferior error. The deviation was calculated in the axial, sagittal, and coronal planes to generate the magnitude of euclidean error for each animal (total n = 180).

Statistical Analysis

Three insertions did not have their corresponding planning scans saved and were excluded. Calculations were performed 4 times, with each trajectory being analyzed by 2 independent operators who repeated their calculations 3 months apart to ensure reliability.18,20 Linear statistics were used to describe the mean and standard deviations of total euclidean error magnitudes. Spherical statistics were used to assess the positional distribution of cannula placements relative to the intended target.6,13 Statistical analysis was performed using built-in functions of IBM SPSS Statistics (IBM, Inc.) and MATLAB R2018a (MathWorks) software. Specifically, the Alpha Model of Reliability Analysis and the Bivariate Correlate subscales in SPSS were used to calculate intra- or interclass correlation and Pearson’s correlation coefficients, respectively. T-test (a,b) and plot3 (x,y,z) MATLAB codes were used to perform paired t-tests and graph euclidean error trajectories, respectively.

Results

Fifteen cannula insertions targeting the putamen were performed with the occipitally placed BJGA in 9 adult rhesus monkeys. The calculated mean euclidean error magnitude was 1.18 ± 0.60 mm (Fig. 4). Directional analyses were performed with spherical statistics, because linear statistics are not appropriate in evaluating the positional distribution of vectors,13 and the elevation and azimuth angles were determined. The elevation angle was found to be 25.5° and the azimuth angle was −75.0°. These angles indicate that the average cannula tip landed posteriorly, superiorly, and rightward of the planned trajectory tip (Supplementary Fig. 1).

FIG. 4.
FIG. 4.

Error magnitude analysis of the BJGA. A box-and-whisker plot of 180 error vector magnitudes shows a normal distribution (median = 1.10 mm, mean = 1.18 mm), with a minimum-maximum range of 0.06 mm and 2.58 mm, respectively, an interquartile range of 0.72–1.58 mm, and an SD of ± 0.60 mm.

Intraclass correlation coefficients within each reviewer’s analyses were 0.89 and 0.96, and the interclass correlation coefficient between their complete analysis was 0.92; this confirmed reliability in methodology of error analysis. Individual reviewer analyses can be seen in Supplementary Figs. 2 and 3.

Discussion

Real-time CED allows for the safe and reliable delivery of therapeutic agents to specific brain regions. To increase the adoption and feasibility of preclinical NHP stereotactic investigations we designed the BJGA, a novel frameless stereotactic guidance device. After multiple intracerebral insertions with the BJGA, the calculated mean euclidean error was 1.18 ± 0.60 mm, with a tendency for the catheter tip to land posteriorly, superiorly, and rightward of the planned target point. Thus, the BJGA demonstrates similar levels of accuracy to currently available frameless options, in addition to being smaller in size, easier to manipulate, and compatible with most onboard MRI software. We have used the BJGA to target various structures within the NHP brain, including the thalamus, caudate, entorhinal cortex, and lateral ventricle.27,28,36

Previous analyses of the accuracy of frameless systems have used two definitions to assess platform efficacy (Supplementary Fig. 4). Euclidean error is a 3D distance between the planned target point and the actual cannula tip. In contrast, the radial error is a 2D distance between the intended and actual cannula locations, measured in a plane perpendicular to the target.31 Radial error is commonly reported when placing deep brain stimulation electrodes, because radial deviations of the electrode cannot be overcome by the activation of different contact leads. Euclidean error is more relevant to parenchymal infusions, because small cannula tip deviations may significantly impact total coverage of the target structure.42,43

We demonstrate that the BJGA allows for a comparable degree of accuracy to the Nexframe and ClearPoint systems. Reported euclidean error values for the Nexframe system range from 1.0 ± 0.8 mm25,26 to 2.18 ± 0.92 mm.39 The discrepancy in calculated mean euclidean errors between these studies may be attributed to differences in imaging protocols or experience of the surgical team.16 Initial validation studies of the ClearPoint system in the NHP brain reported a mean euclidean error of 0.8 ± 0.2 mm,32 whereas subsequent studies in 6 human patients8 reported a mean euclidean error of 1.9 ± 0.9 mm. The directional deviations observed with the BJGA are consistent with those seen in the ClearPoint system, which were posited to be due to nonlinearities in MR space.22

The larger dimensions of the Nexframe system are not conducive to multiple ipsilateral or simultaneous bilateral infusions in humans and NHPs. The smaller SmartFrame allows for simultaneous bilateral cannula insertions in humans, but not in NHPs. In contrast, the small size of the BJGA allows for multiple simultaneous ipsilateral and contralateral implantations, enabling synchronized multitarget infusions, which could significantly reduce overall surgical and anesthesia time. The small size of the BJGA allows for prone and lateral patient positioning within the MR bore and enables surgeons to readily consider and use frontal, parietal, temporal, and occipital entry points in their planning, allowing them to plan trajectories that were not previously feasible. Furthermore, the close proximity of parallel tracts within the guide allows for “rescue infusions,” or the ability to slightly adjust an infusion trajectory to correct what is observed to be a suboptimal infusion. Finally, the relatively small and light profile of the device may allow future applicability to human pediatric neurosurgical cases. Table 1 compares aspects of the BJGA to other available frameless systems.

TABLE 1.

Summary of the key features of BJGA compared to several available frameless platforms

FeatureBJGAClearPoint SmartFrame (MRI Interventions)Nexframe (Medtronic)Axis (Monteris)Navigus (Medtronic)
MRI compatibleX
Allows for trajectory adjustments w/ in the MR scannerXXX
Compatible w/ any MR softwareXXXX
Capable of multiple implantationsXXXX
No. of access trajectories from a single frame alignment31111
Dimensions—height × width (cm)3.2 × 2.212 × 510.8 × 6.010 × 135.5 × 2.5

An important advantage of the BJGA is that it does not require rigid head fixation within the MR scanner. The ClearPoint system defines intracranial coordinates according to the isocenter of the MR scanner, and any head movement alters the target coordinates and necessitates a replanning step. In contrast, BJGA trajectories are defined relative to the intracranial target and the device position (astride the burr hole), and are independent of head position. The BJGA also allows surgeons to make trajectory adjustments within the MR scanner. This is important for accurately targeting brain nuclei, because brain shift has been described following minimally invasive burr hole procedures.15 Thus, adjustments based on iMRI allow for improved accuracy of planned trajectories.

Another important simplification of the BJGA is its universal MRI console adaptability. Current guiding devices require sophisticated neuronavigation software. The Nexframe system has a complex onboard neuronavigation software that requires specific training and a clear view of the surgical field for registration of independent spherical fiducials (which can be subject to optical imperfections).31 The ClearPoint system requires the use of proprietary navigation and planning software in the MRI suite, as well as technical assistance from a qualified technician. With the NHP version of the BJGA, trajectories can be planned accurately with various MR scanners and their onboard software (unpublished data), making the BJGA easily adaptable to current and future software. The NHP version of the BJGA does not require preplanning software, due to easily defined landmarks. We are developing a clinical version of the BJGA that can be used concomitantly with image-guided surgical navigation systems, such as the StealthStation (Medtronic, Inc.), and we will describe the clinical device and its integration with neuronavigation systems in a future manuscript. We anticipate that the universal adaptability of the NHP BJGA will help reduce acquisition and per-case costs—making preclinical adoption more attractive.

Poor catheter positioning has been linked to the limited efficacy observed in previous brain infusion trials for Parkinson disease5 and glioblastoma.37 When combined with recent improvements in CED, iMRI guidance, and specifically designed intracerebral cannulae,3,7,12,33,40 the improved targeting of deep brain structures afforded by the BJGA has the potential to result in substantial improvements in the safety, accuracy, and efficacy of neurosurgically delivered investigational therapeutic agents.

In summary, the BJGA is an attractive option for preclinical gene therapy and other experiments in NHPs requiring precision guidance, due to its comparable accuracy to currently available devices, smaller and lighter profile, ease of use, wide applicability, and lower overall cost. Although this study primarily assessed the accuracy of the BJGA for cannula tip placement for infusion within a specific locus, we anticipate that similar technologies and methods can be modified for use in additional neurosurgical interventions, including but not limited to laser ablations, biopsies, targeted lesioning, injections, device implantations, and microdialyses in NHPs. Following additional validation studies in NHPs and phantom models, we anticipate that the BJGA will become an alternative image-guided targeting platform in the neurosurgical armamentarium.

Conclusions

In this study we discuss the development of a novel targeting device for use in preclinical iMRI gene therapy procedures in NHPs. The BJGA provides a degree of accuracy comparable to currently available frameless stereotactic options, in addition to being smaller, lighter, and compatible with nearly all onboard MR scanning software. Given these features, the BJGA provides an accurate and user-friendly option for preclinical image-guided neurosurgery experiments.

Acknowledgments

This project was supported by grants from the NIH (Grant No. P01 CA118816) and Kinetics Foundation to Dr. Bankiewicz.

Disclosures

Drs. Bankiewicz and Kells report that they are inventors on a patent (WO2018044933A1) describing the BJGA. Dr. Kells was an employee of Voyager Therapeutics, a publicly traded company, and has direct stock ownership in that company. Drs. Kells and Fiandaca, as current employees of Brain Neurotherapy Bio, Inc., a private company, have direct stock ownership in that company.

Author Contributions

Conception and design: Bankiewicz. Acquisition of data: Bankiewicz, Sudhakar, Bringas. Analysis and interpretation of data: Bankiewicz, Sudhakar, Mahmood. Drafting the article: Sudhakar, Naidoo. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Bankiewicz. Statistical analysis: Mahmoodi. Administrative/technical/material support: Bankiewicz. Study supervision: Bankiewicz.

Supplemental Information

Online-Only Content

Supplemental material is available with the online version of the article.

Current Affiliations

Drs. Kells and Fiandaca: Brain Neurotherapy Bio, Inc., Oakland, CA.

References

  • 1

    Attenello FJ, Mukherjee D, Datoo G, McGirt MJ, Bohan E, Weingart JD, et al.: Use of Gliadel (BCNU) wafer in the surgical treatment of malignant glioma: a 10-year institutional experience. Ann Surg Oncol 15:28872893, 2008

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Bakay RA, Fiandaca MS, Sweeney KM, Colbassani HJ Jr, Collins DC: Delayed stereotactic transplantation technique in non-human primates. Prog Brain Res 78:463471, 1988

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Bankiewicz KS, Eberling JL, Kohutnicka M, Jagust W, Pivirotto P, Bringas J, et al.: Convection-enhanced delivery of AAV vector in parkinsonian monkeys; in vivo detection of gene expression and restoration of dopaminergic function using pro-drug approach. Exp Neurol 164:214, 2000

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Bankiewicz KS, Sudhakar V, Samaranch L, San Sebastian W, Bringas J, Forsayeth J: AAV viral vector delivery to the brain by shape-conforming MR-guided infusions. J Control Release 240:434442, 2016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Bartus RT, Herzog CD, Chu Y, Wilson A, Brown L, Siffert J, et al.: Bioactivity of AAV2-neurturin gene therapy (CERE-120): differences between Parkinson’s disease and nonhuman primate brains. Mov Disord 26:2736, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Berens P: CircStat: A MATLAB toolbox for circular statistics. J Stat Softw 31:v031.i10, 2009

  • 7

    Bobo RH, Laske DW, Akbasak A, Morrison PF, Dedrick RL, Oldfield EH: Convection-enhanced delivery of macromolecules in the brain. Proc Natl Acad Sci U S A 91:20762080, 1994

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Chittiboina P, Heiss JD, Lonser RR: Accuracy of direct magnetic resonance imaging-guided placement of drug infusion cannulae. J Neurosurg 122:11731179, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Fiandaca MS, Bankiewicz KS: Gene therapy for Parkinson’s disease: from non-human primates to humans. Curr Opin Mol Ther 12:519529, 2010

  • 10

    Fiandaca MS, Forsayeth JR, Dickinson PJ, Bankiewicz KS: Image-guided convection-enhanced delivery platform in the treatment of neurological diseases. Neurotherapeutics 5:123127, 2008

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Fiandaca MS, Salegio EA, Yin D, Richardson RM, Valles FE, Larson PS, et al.: Human/nonhuman primate AC-PC ratio—considerations for translational brain measurements. J Neurosci Methods 196:124130, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Fiandaca MS, Varenika V, Eberling J, McKnight T, Bringas J, Pivirotto P, et al.: Real-time MR imaging of adeno-associated viral vector delivery to the primate brain. Neuroimage 47 (Suppl 2):T27T35, 2009

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Fisher NI: Statistical Analysis of Circular Data. New York: Cambridge University Press, 1993

  • 14

    Gorgulho AA, Shields DC, Malkasian D, Behnke E, Desalles AAF: Stereotactic coordinates associated with facial musculature contraction during high-frequency stimulation of the subthalamic nucleus. J Neurosurg 110:13171321, 2009

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Ivan ME, Yarlagadda J, Saxena AP, Martin AJ, Starr PA, Sootsman WK, et al.: Brain shift during bur hole-based procedures using interventional MRI. J Neurosurg 121:149160, 2014

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Kelman C, Ramakrishnan V, Davies A, Holloway K: Analysis of stereotactic accuracy of the Cosman-Robert-Wells frame and Nexframe frameless systems in deep brain stimulation surgery. Stereotact Funct Neurosurg 88:288295, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 17

    Kitchen ND, Lemieux L, Thomas DG: Accuracy in frame-based and frameless stereotaxy. Stereotact Funct Neurosurg 61:195206, 1993

  • 18

    Koch GG: Intraclass correlation coefficient, in Kotz S, Read CB, Balakrishnan N, et al. (eds): Encyclopedia of Statistical Sciences. Hoboken, NJ: Wiley, 2006 (https://onlinelibrary.wiley.com/doi/full/10.1002/0471667196.ess1275.pub2) [Accessed November 16, 2018]

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Konrad PE, Neimat JS, Yu H, Kao CC, Remple MS, D’Haese PF, et al.: Customized, miniature rapid-prototype stereotactic frames for use in deep brain stimulator surgery: initial clinical methodology and experience from 263 patients from 2002 to 2008. Stereotact Funct Neurosurg 89:3441, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Koo TK, Li MY: A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 15:155163, 2016

  • 21

    Kordower JH, Fiandaca MS: Response of the monkey cholinergic septohippocampal system to fornix transection: a histochemical and cytochemical analysis. J Comp Neurol 298:443457, 1990

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22

    Larson PS, Starr PA, Bates G, Tansey L, Richardson RM, Martin AJ: An optimized system for interventional magnetic resonance imaging-guided stereotactic surgery: preliminary evaluation of targeting accuracy. Neurosurgery 70 (1 Suppl Operative):95103, 2012

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Leung CH, Kliem MA, Heeke BL, McPhee SWJ, Federici T, Snyder BR, et al.: Assessment of hippocampal adeno-associated viral vector gene delivery via frameless stereotaxis in a nonhuman primate. Stereotact Funct Neurosurg 89:275285, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Lu Y, Yeung C, Radmanesh A, Wiemann R, Black PM, Golby AJ: Comparative effectiveness of frame-based, frameless, and intraoperative magnetic resonance imaging-guided brain biopsy techniques. World Neurosurg 83:261268, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    Martin AJ, Larson PS, Ostrem JL, Keith Sootsman W, Talke P, Weber OM, et al.: Placement of deep brain stimulator electrodes using real-time high-field interventional magnetic resonance imaging. Magn Reson Med 54:11071114, 2005

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26

    Martin AJ, Larson PS, Ostrem JL, Starr PA: Interventional magnetic resonance guidance of deep brain stimulator implantation for Parkinson disease. Top Magn Reson Imaging 19:213221, 2009

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Nagahara AH, Wilson BR, Ivasyk I, Kovacs I, Rawalji S, Bringas JR, et al.: MR-guided delivery of AAV2-BDNF into the entorhinal cortex of non-human primates. Gene Ther 25:104114, 2018

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 28

    Naidoo J, Stanek LM, Ohno K, Trewman S, Samaranch L, Hadaczek P, et al.: Extensive transduction and enhanced spread of a modified AAV2 capsid in the non-human primate CNS. Mol Ther 26:24082430, 2018

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29

    Neuwelt EA: Mechanisms of disease: the blood-brain barrier. Neurosurgery 54:131142, 2004

  • 30

    Oldendorf WH: Lipid solubility and drug penetration of the blood brain barrier. Proc Soc Exp Biol Med 147:813815, 1974

  • 31

    Palys V, Holloway KL: Frameless functional stereotactic approaches. Prog Neurol Surg 33:168186, 2018

  • 32

    Richardson RM, Kells AP, Martin AJ, Larson PS, Starr PA, Piferi PG, et al.: Novel platform for MRI-guided convection-enhanced delivery of therapeutics: preclinical validation in nonhuman primate brain. Stereotact Funct Neurosurg 89:141151, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 33

    Richardson RM, Kells AP, Rosenbluth KH, Salegio EA, Fiandaca MS, Larson PS, et al.: Interventional MRI-guided putaminal delivery of AAV2-GDNF for a planned clinical trial in Parkinson’s disease. Mol Ther 19:10481057, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 34

    Richardson RM, Varenika V, Forsayeth JR, Bankiewicz KS: Future applications: gene therapy. Neurosurg Clin N Am 20:205210, 2009

  • 35

    Salegio EA, Kells AP, Richardson RM, Hadaczek P, Forsayeth J, Bringas J, et al.: Magnetic resonance imaging-guided delivery of adeno-associated virus type 2 to the primate brain for the treatment of lysosomal storage disorders. Hum Gene Ther 21:10931103, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 36

    Samaranch L, Blits B, San Sebastian W, Hadaczek P, Bringas J, Sudhakar V, et al.: MR-guided parenchymal delivery of adeno-associated viral vector serotype 5 in non-human primate brain. Gene Ther 24:253261, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 37

    Sampson JH, Archer G, Pedain C, Wembacher-Schröder E, Westphal M, Kunwar S, et al.: Poor drug distribution as a possible explanation for the results of the PRECISE trial. J Neurosurg 113:301309, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 38

    San Sebastian W, Richardson RM, Kells AP, Lamarre C, Bringas J, Pivirotto P, et al.: Safety and tolerability of magnetic resonance imaging-guided convection-enhanced delivery of AAV2-hAADC with a novel delivery platform in nonhuman primate striatum. Hum Gene Ther 23:210217, 2012

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 39

    Starr PA, Martin AJ, Ostrem JL, Talke P, Levesque N, Larson PS: Subthalamic nucleus deep brain stimulator placement using high-field interventional magnetic resonance imaging and a skull-mounted aiming device: technique and application accuracy. J Neurosurg 112:479490, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 40

    Valles F, Fiandaca MS, Eberling JL, Starr PA, Larson PS, Christine CW, et al.: Qualitative imaging of adeno-associated virus serotype 2-human aromatic L-amino acid decarboxylase gene therapy in a phase I study for the treatment of Parkinson disease. Neurosurgery 67:13771385, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 41

    Yin D, Fiandaca MS, Forsayeth J, Bankiewicz KS: Real-time convection delivery of therapeutics to the primate brain, in Janowski M (ed): Experimental Neurosurgery in Animal Models. New York: Springer, 2016, Vol. 116, pp 175194

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 42

    Yin D, Richardson RM, Fiandaca MS, Bringas J, Forsayeth J, Berger MS, et al.: Cannula placement for effective convection-enhanced delivery in the nonhuman primate thalamus and brainstem: implications for clinical delivery of therapeutics. J Neurosurg 113:240248, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 43

    Yin D, Valles FE, Fiandaca MS, Bringas J, Gimenez F, Berger MS, et al.: Optimal region of the putamen for image-guided convection-enhanced delivery of therapeutics in human and non-human primates. Neuroimage 54 (Suppl 1):S196S203, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation

Supplementary Materials

  • Collapse
  • Expand

Image from Ryu et al. (pp 442–455).

  • FIG. 1.

    Schematics of the BJGA: measurements, implantation landmarks, and device assembly. A: Main parts and dimensions of the BJGA. The 3-hole ball array with adjustable side screw (black dot) is the core of the device, and guides the cannula to the targeted structure (1). The threaded skull-mounted base is secured onto the skull by 3 titanium fasteners (2). The knurled threaded locking collar interlocks with the base and secures the device in the desired position (3). B: Apical view of the ball array and the skull-mounted base with the most relevant measurements. C: Assembly of the BJGA with all parts. D: Another essential piece of the BJGA is the gadolinium-based fluid–filled guiding tubing that is placed inside the ball array during trajectory planning (1). As depicted in the cross-section diagram (2), this fiducial consists of a threaded cap (red arrowheads), a small chamber (blue arrowheads), and 3 PEEK tubes (green arrowheads) filled with a 2-mM solution of a gadolinium-based MR contrast agent. E: Representative T1W MR image of the BJGA and the fiducial tubes, demonstrating the gadolinium signal from the chamber and the tubing. The BJGA is not visible on MRI and has been contoured with dashed lines. Red arrowheads represent the threaded cap, blue arrowheads represent the gadolinium-filled chamber, and green arrowheads represent the PEEK tubes. F: Lateral view of the BJGA on an NHP skull model with the fiducials inserted in the ball. The maximum angulation of 16° is shown in red. G: Superior view of bilateral implantation of the BJGA on an NHP skull model with the fiducials inserted in the ball. The maximum angulation of 16° is shown in red. H: Posterior view of the BJGA with (right) and without (left) the locking collar. The maximum angulation of 16° is shown in red. I: Posterior view of an NHP skull with the skull-mounted base placed bilaterally. Relevant anatomical landmarks on the NHP skull model—the midline (black arrowheads) and the nuchal ridge (white arrowheads), are noted.

  • FIG. 2.

    Workflow for the digital alignment of the fiducial’s gadolinium signal of the guiding device using the MPR toolbox. Before target selection and trajectory planning, the xyz planes must be oriented such that the entire fiducial length is visible in the MRI sequences. First, T1W MRI sequences were acquired, and a portion of the gadolinium-positive fiducial tubes was visualized (white arrowhead). Step 1: The x plane is oriented on top of the midline (axial); the z plane is oriented parallel to the AC-PC line (sagittal). Step 2: The crosshair (white circle) is placed on top of one of the fiducial signals (straight arrow), and the z plane is oriented to the fiducial signal (sagittal; curved arrow). Step 3: The x plane is rotated to align with the gadolinium signal in the axial plane (curved arrow). Small refinements can be done in the sagittal plane to correctly align the z plane to the fiducial signal. Step 4: The crosshair is located in the center of the fiducial tube in the coronal view. Numbers 1–3 label each of the 3 holes of the ball array.

  • FIG. 3.

    Schematic workflow for appropriate BJGA manipulation to the selected target (i.e., putamen). Once the BJGA is mounted and secured on the skull, the array can be manually adjusted to target the selected structure. Initially, if the array is not targeting the structure (A, white arrowheads depict missing putamen) multiple adjustments in the medial-lateral and superior-inferior direction can be made until the trajectory targets the structure (B [left panel, white arrowhead], shows the trajectory medial to the putamen, and B [right panel, white arrowhead], shows the trajectory appropriately targeting the putamen). Once the trajectory is defined (C), the surgeon selects the appropriate access hole (D). The surgeon must then ensure that the trajectory is not proximal to sulci or ventricles, to avoid a potential large-vessel insult. If the trajectory is not satisfactory, the surgeon can manually adjust the BJGA to redefine the trajectory. Image (E) shows an example of suboptimal trajectory that crosses too close to a sulcus (white arrowhead). Image (F) shows how the previous trajectory is altered by superiorly adjusting the BJGA, and changing the entry hole from lateral to medial (insets, panels E and F). The new trajectory satisfies the safety check with regard to the sulcus (white arrowhead). The surgeon then measures the depth from the bottom of the gadolinium-filled chamber (white arrow) to the target point (i.e., putamen), subtracts 3 mm to account for the distance between the bottom of the chamber and the top of the BJGA (see Fig. 1D and 1E), and marks the distance on the cannula. A depth stop is placed at the target mark, the cannula is introduced through the selected entry hole to the target, and the cannula location is determined (G). If the cannula tip is in an optimal position, the infusion can begin (H).

  • FIG. 4.

    Error magnitude analysis of the BJGA. A box-and-whisker plot of 180 error vector magnitudes shows a normal distribution (median = 1.10 mm, mean = 1.18 mm), with a minimum-maximum range of 0.06 mm and 2.58 mm, respectively, an interquartile range of 0.72–1.58 mm, and an SD of ± 0.60 mm.

Metrics

All Time Past Year Past 30 Days
Abstract Views 1850 135 0
Full Text Views 584 341 58
PDF Downloads 363 89 8
EPUB Downloads 0 0 0