Niranjan A, Lunsford LD, Richardson RM (eds): Current Concepts in Movement Disorder Management. Prog Neurol Surg. Basel, Karger, 2018, vol 33, pp 168–186 (DOI: 10.1159/000481102)
______________________
Frameless Functional Stereotactic Approaches
Viktoras Palysa, b · Kathryn L. Hollowaya, b
aDepartment of Neurosurgery, Medical College of Virginia Hospitals, Virginia Commonwealth University, and bThe Parkinson’s Disease Research, Education and Care Center, Hunter Holmes McGuire Department of Veterans Affairs Medical Center, Richmond, VA, USA
______________________
Abstract
The stereotactic frame has served as the gold standard apparatus for accurate and precise targeting of deep brain structures since 1947. Despite passing the test of time, the stereotactic frame has several limitations from the perspective of both neurosurgeons and patients. Therefore, there was a need to develop a frameless system that had equivalent accuracy and reliability to the frame. This need was met with 3 commercially available frameless stereotactic systems designed specifically for deep brain stimulation surgery: Nexframe, STarFix, and ClearPoint. Over the past decade, the frameless and frame-based systems have been extensively investigated by numerous studies and found to be equivalent in experimental and clinical accuracy as well as in clinical outcomes. This chapter summarizes the findings of those studies along with the discussion of sources of stereotactic errors. The procedural aspects, advantages, and disadvantages of each frameless system are reviewed. Frameless stereotaxy is a safe, accurate, and effective technique for functional stereotactic approaches and provides a viable alternative to the frame-based systems.
© 2018 S. Karger AG, Basel
Introduction
Fundamental to the principle of stereotaxy is a surgical instrument guide that affixes rigidly, aligns accurately, and guides instruments precisely; hand-held methods are ipso facto not stereotactic [1]. The stereotactic frame fulfills those criteria. Since 1947, when Spiegel et al. [2] introduced the first human stereotactic frame for neurosurgical procedures, it has served as the gold standard apparatus for accurate and precise targeting of deep brain structures [3–5]. There have been multiple modifications of the stereotactic frame over time, with the majority of modern deep brain stimulation (DBS) procedures performed with stereotactic arc systems – either the Cosman-Roberts-Wells (CRW) or Leksell frames. Despite passing the test of time, the stereotactic frame has several limitations.
From the patient’s perspective – it is a restrictive and heavy device. The stereotactic frame is placed on the morning of surgery, and the stereotactic images are then acquired, requiring the patient to be transported back and forth between the operating room and the scanner. During the long operative day, the stereotactic frame is rigidly attached to the operating table, while the awake patient is asked to actively participate in testing. The patient looks out at the examiner through bars with the head held in a vise grip, which can generate feelings of anxiety. There are patients who have refused to have surgery because they fear the application and hours spent in the constraining frame.
From the surgeon’s perspective – placement of the stereotactic frame and patient positioning can be challenging in patients with very large or very small head diameters, cervical/thoracic kyphosis or claustrophobia. Patient motion due to tremor or dystonia can result in slippage of the frame endangering the safety of the surgery. Additionally, the frame limits airway access. Lastly, the frame serves as a barrier to easy observation of the patient by the examiner and vice versa. This can result in less than optimal detection of stimulation-induced side effects in the face [6, 7].
With the advent of modern neuronavigation technology, it became possible to perform stereotactic neurosurgery without the conventional frame [8, 9]. For procedures that require a stereotactic aiming device, there are 2 types of skull-mounted frameless designs in current use: (A) bur hole-seated instrument guides based on a trapped ball; and (B) skull-affixed platforms (miniframes). DBS surgery has additional demands beyond all other applications of stereotaxy, such as brain biopsy needle, depth electrode or laser probe placements. It requires the greatest accuracy and a stable ergonomic platform for microelectrode recording and stimulation over the course of the operation and often through multiple precisely parallel trajectories.
Therefore, there was a need to develop a stereotactic frameless system, which had equivalent accuracy and reliability to the traditional stereotactic frame with the benefits of a frameless system. This need was met with 3 commercially available frameless stereotactic systems, designed specifically for DBS surgery:
1 Nexframe® (Medtronic, Inc., Minneapolis, MN, USA) – Food and Drug Administration (FDA) approved in 2001;
2 STarFix™ (FHC, Inc., Bowdoin, ME, USA) – FDA approved in 2001;
3 ClearPoint® (MRI Interventions, Inc., Irvine, CA, USA) – FDA approved in 2010.
Accuracy of Stereotactic Systems, Targeting Errors
Stereotactic accuracy directly relates to DBS efficacy [10–13]. For example, the subthalamic nucleus (STN) is approximately 6 × 4 × 5 mm in size and a DBS electrode placement error of as little as 2 mm can affect the outcome [13–15]. Thus, the accuracy and clinical efficacy of stereotactic systems have been extensively investigated and compared.
The targeting accuracy of stereotactic system depends on the contribution (vector summation) of 2 autonomous sources of error:
1 Hardware mechanical accuracy (also known as phantom accuracy) – determined by characteristics of platform, such as intrinsic stability, registration error, and ability to accurately align to a distant target. It can be evaluated in isolation in phantom experiments and is frequently advertised by device manufacturers.
2 Accuracy of imaging modalities and parameters chosen (esp. scan slice thickness).
Clinical studies investigating the application targeting accuracy (also known as clinical, real-world, final or total accuracy) demonstrate greater error for both frame and frameless systems, as compared with phantom experiments. This is expected due to several additional sources of error:
1 The effect of positioning during the scan versus surgery. Rohlfing et al. [16] found a decrease in the accuracy of stereotactic frames by the effects of the mechanical loading on the frame and a change in position from supine to prone. A smaller effect can be assumed when patients are imaged flat, but operated on in a semi-recumbent position.
2 The brain shift due to the normal mobility within the cranial cavity as well as brain sag induced by the loss of cerebrospinal fluid. The importance of this phenomenon was demonstrated in the studies by D’Haese et al. [17] and Ivan et al. [18].
3 The inaccuracy of localization of the DBS electrode contacts on the imaging. Both MRI magnetic susceptibility artifact and CT beam hardening artifact conceal the precise electrode center and require the estimation or interpolation of the contact position, thus, introducing the localization errors along with CT-MRI fusion errors. Plus, each modality – MRI and CT – has its own inherent inaccuracies with corresponding measurement errors. Thani et al. [19] evaluated 8 consecutive patients with DBS electrodes and found the coregistration discrepancy between intraoperative MRI and postoperative CT to be 1.6 ± 0.2 or 1.5 ± 0.2 mm, depending on the software used. Holloway and Docef [20] demonstrated that repeated CT imaging of stationary bone fiducials resulted in a localization discrepancy of 0.7 mm (range 0–1.8 mm). Direct measurement of localization errors in phantom studies eliminates this step of translating postoperative imaging to preoperative targeting.
4 The technical errors in lead placement. Examples include deflection of the implanted device by the dura or bone edge and slippage of the DBS lead during or after anchoring.
It is also important for the reader to understand how hardware-positioning errors are reported:
1 The vector error (sometimes called Euclidian error, tip error or total error) is the 3D (Euclidian, xyz) vector length between the intended target and actual DBS lead tip location. The vector error may not always be clinically relevant, as small lead misplacements (≤2 mm) along the axis of implantation (z) can easily be corrected by the selection of a different active contact from the quadripolar lead.
2 The radial error (r) is the 2D (xy) distance between the actual lead and the intended target measured on the axial imaging plane of the target. This type of error is commonly tested in phantom studies using a targeting disc or in clinical studies in an effort to dismiss the clinically less relevant differences in depth along the axis of implantation (z).
3 The trajectory error (t) (sometimes called lateral error) – the closest distance between the actual lead trajectory and the intended trajectory at target level. If the intended and actual trajectories are parallel, trajectory error is similar to the radial error. This becomes more relevant when the trajectories are not parallel, as this error will affect the utility of the rest of the contacts.
Phantom Accuracy Studies
The accuracy ...