Robotics: Background and Current Role

8 Robotics: Background and Current Role

Sean M. Barber, Meng Huang, and Paul Holman, and Isador H. Lieberman

Summary

Early experience with spinal robotic systems have shown them to be a useful tool for assisting with pedicle screw placement, albeit with some limitations. Future improvements in robotic technology have the potential to expand the applications of these devices to involvement with more advanced spinal surgery tasks.

Keywords: spinal robotics robotic surgery pedicle screw spinal navigation

8.1 Introduction

In the field of surgery, improvements in diagnostic and therapeutic capability are often driven by technological innovation. Advances in sterile technique (e.g., autoclave), hemostasis (e.g., bipolar electrocautery), diagnostic imaging (e.g., computed tomography [CT], magnetic resonance imaging), and surgical visualization (e.g., operating microscope, endoscopy) have each led to improvements in surgical potential and/or outcomes, together drastically altering the landscape of surgical practice as a whole. Not every advance in surgical technology, however, proves useful or cost-effective enough to warrant widespread adoption. Occasionally, technological progress arises without identifiable applications or with flaws that dilute its value and practicality to unacceptable levels. Rigorous and objective analysis of new surgical technology is paramount to ensuring that the benefits provided are not outweighed by the costs of implementation, and that technology for the sake of technology is rejected.

Over the past decade, stereotactic spinal navigation has slowly emerged as a widespread enabling technology in both complex and minimally invasive spinal surgery. Value has been demonstrated in improved accuracy of screw insertion, reduction in occupational radiation exposure, and facilitation of both minimally invasive and complex revision surgery. Robotic surgical systems designed specifically for spine surgery are now emerging as competitive or complementary tools for the use of spinal navigation. Although robotic surgery has proven to be a feasible alternative to open and endoscopic/laparoscopic techniques in other surgical specialties,1^,²^,³^,⁴^,⁵^,⁶^,⁷ robotic-assisted spinal surgery has yet to achieve widespread adoption. The role of robotic devices in the everyday spine surgery practice remains ill-defined. What benefits do such devices confer? How can they be effectively incorporated into the operative workflow? How steep is the learning curve, and what potential risks are involved? The goal of this chapter is to address these concerns and provide an understanding of the current state and future directions of robotics in spine surgery.

8.2 Benefits and Limitations of Robotics in Spine Surgery

A wealth of potential opportunities exists for the incorporation of robotics into spine surgery, and the theoretical benefits associated with robotic surgical assistance in surgery in general—and spine surgery in particular—are numerous. The benefits of robotic surgical assistance can be divided into two broad categories: (1) reduction of human error, thus improving the safety and efficacy of current practice, and (2) augmenting human surgical capability such that procedures, approaches, and treatments, which were previously impractical are now feasible, reproducible, and safe.

Human error plays a role in a number of intraoperative complications in spine surgery. Humans are subject to fatigability, distraction, tremor, inconsistency and relative limitations in dexterity, memory, and an ability to detect and anticipate changes in the environment and surgical field, many of which may be improved upon with robotic assistance.

However, modern robotic technology also bears certain weaknesses including limited dexterity, high monetary cost, relatively large operating room space requirement, and poor independent decision-making capabilities, among others.

Spine surgery in particular seems well-suited to the use of robots, as many modern spinal procedures involve repeatedly performing similar mechanical tasks of relatively low complexity at multiple levels (e.g., laminectomy, discectomy, pedicle screw placement). Spinal robotic technology is still in its infancy, however, and the majority of objective evidence available regarding the benefits of robotic devices in spine surgery draws from the use of robots in a shared control model to assist with the placement of pedicle screws.

8.2.1 Pedicle Screw Accuracy

Currently available spinal robotic systems are passive and primarily assist with the localization and execution of a preplanned trajectory, lending themselves well to the placement of pedicle screws. There is little consensus, however, regarding how the accuracy of robot-assisted screws compare with screws placed via other methods. Several studies have shown significantly improved accuracy of pedicle screws placed with robotic assistance compared with conventional, fluoroscopy-guided or navigated techniques,¹⁴^,¹⁵^,¹⁷ whereas others have shown statistically comparable⁸^,¹⁰^,¹²^,¹³^,¹⁶^,¹⁸ or inferior results⁹ with robotic assistance. Many confounding variables exist when making such comparisons, however, including surgeon experience, the specific robotic device used, the method of intraoperative registration employed, whether the screws were placed percutaneously or open, and whether the robot was affixed directly to the bony anatomy (e.g., via a spinous process clamp or Schanz screw) or instead attached only to the operating table.

Roser et al evaluated 37 patients randomized to undergo pedicle screw instrumentation with freehand techniques (N = 10), standard navigation (N = 9), and the Mazor™ SpineAssist (Mazor Robotics Ltd., Caesarea, Israel) (N = 18).¹⁹ Although they found a trend toward improved accuracy in the robotic group (99% Gertzbein and Robbins [GRS] A) compared with the navigated (92% GRS A) and freehand groups (97.5% GRS A), no statistical analysis was performed due to small sample sizes.

Schröder et al performed a literature review of 14 articles comparing the need for intraoperative screw revision or revision surgery for screw malposition in patients undergoing pedicle screw placement with robotic guidance, conventional freehand techniques, or CT/fluoroscopic navigation. They found that intraoperative screw revisions were significantly higher for navigated screws when compared with robot-guided screws (p < 0.001, odds ratio [OR] 9.7, 95% confidence interval [CI] 2.9–32.2) and revision surgery for screw malposition was significantly higher for freehand compared with robot-guided screws (p < 0.001, OR 8.1, 95% CI 2–33.3).²⁷

Most modern robotic spine surgical devices utilize a preoperative CT scan or an intraoperative conebeam CT data set when planning screw trajectories. Little objective evidence exists regarding the superiority of using either a preoperative or intraoperative CT scan for planning. It is interesting to note, however, that Macke et al, after retrospectively evaluating pediatric patients who underwent robot-assisted placement of 662 pedicle screws, found that screw accuracy was significantly higher when the planning CT was taken in the prone position compared with the supine position.¹¹

8.2.2 Radiation Exposure

Spinal robots could also theoretically decrease radiation exposure to the patient and surgical team during certain procedures involving spinal instrumentation. Indeed, some series have shown decreased radiation exposure to the surgeon and lower fluoroscopy times per screw when placing pedicle screws with robotic-assistance compared with fluoroscopic guidance,14^,¹⁵^,¹⁶^,¹⁹^,²¹ though other studies have failed to confirm this.⁹^,¹⁰^,¹² Solomiichuk et al retrospectively reviewed 70 patients (406 screws) with metastatic spinal disease requiring thoracolumbar instrumentation with the SpineAssist robot (N = 35, 192 screws) or conventional, fluoroscopic guidance (N = 35, 214 screws) and found that while overall radiation time requirements were comparable between these groups (p = 0.61), radiation intensity was higher in the robotic group (p < 0.01), a finding they attributed to the need for high-quality fluoroscopy required for the robotic system registration process.¹⁸

8.2.3 Operative Times and Blood Loss

Although several series have compared operative times with conventional freehand or fluoroscopy-guided pedicle screw placement to robot-assisted pedicle screws, little consensus exists. A single cadaver series found a significantly decreased operative time for pedicle screw placement compared with conventional fluoroscopy,²¹ while several other series saw no significant difference between robot-assisted and conventional techniques,⁸^,¹⁵^,¹⁶ and still others found a significantly increased operative time with robot-assisted screws.⁹^,¹²^,¹³ One might expect that procedural modifications involving a new device would lead to longer operative times, at least initially. Over time, as the surgeon and operative staff become more accustomed to the new device and workflow, operative times would be expected to improve.

A single series compared blood loss in a mixed group of open and percutaneous robot-assisted pedicle screw procedures to that seen with open conventional fluoroscopy screw placement and found that blood loss was less in the robotic group.⁸

8.2.4 Other Potential Applications

Bony decompression of the neural elements in patients with pain and neurological deficits is arguably the most important task performed during the vast majority of spinal procedures. Achieving an optimal result can be hindered by spatial disorientation particularly in revision cases and during the learning curve of minimally invasive procedures performed with minimal access retractor systems. Planning the site and extent of decompression preoperatively and using passive or semiactive robots with a drill or osteotome as the end effector could help to optimize the amount of the removal of bone, ligament, and disc desired by the surgeon. Rod-bending is another routine task which could theoretically benefit from robotic assistance as rod-bending involves the conformation of an object to fixed parameters such as screw head location. Given that the stereotactic technology for using optical cameras to facilitate the rod-bending process is already available (Bendini, Nuvasive Inc.), it is not difficult to imagine a robotic device performing a similar process and bending/inserting the rod semiautonomously with the goal of achieving optimal sagittal spinal alignment.

Robotic devices also have a tremendous potential benefit in minimally invasive spine procedures. Certain procedural steps that are difficult for humans to perform through small tubes (e.g., dural closure) could theoretically be performed with greater aptitude by a robotic arm. There is also a significant opportunity to pair robotics with endoscopic spinal surgery to improve efficiency and reproducibility to assist surgeons with a relatively steep learning curve.

Finally, although the field of spinal robotics is still relatively rudimentary, further advancements in robotic technology could enhance our current surgical capabilities and allow for new approaches and treatment possibilities that are currently impractical (Table 8.1). Telesurgery, for instance, or the practice of surgery remote from the patient’s location, has been shown to be feasible in other surgical specialties with the use of robotics²²^,²³ and remote surgery or remote assistance/guidance of other surgeons through robotics could theoretically be a possibility in spine surgery.

Table 8.1 Current and potential applications of robotic technology

Screw and implant placement

Bilateral screw placement

Single-stage surgery

Telesurgery/remote surgery

Decompression

Rod-bending

Dural closure

8.2.5 Current and Potential Applications

Table 8.1 lists the applications of the robot-assisted surgery.

8.2.6 Monetary Costs

Modern robot-assisted pedicle screw placement may confer a modest improvement in screw placement accuracy and other desirable parameters, but the devices themselves also impart a considerable economic cost burden, and this cost may limit the availability and use of such devices to high-volume centers. Little objective data exists regarding the cost-effectiveness of such devices, but data from the use of CT navigation in spine surgery suggests that, in high-volume centers and those performing difficult cases, the relatively high initial cost investment is indeed economically justified by a reduction in the need for costly revision surgeries.24^,²⁵ In addition, reduction in the time required for pedicle screw placement relative to conventional methods further compounds cost savings,²⁴ as do reductions in the indirect costs likely associated with undue radiation exposure to the patient, surgeon, and staff.

Furthermore, though the robotic spinal systems currently available are fairly limited in scope and application, the utility of these devices will likely grow with time, and with an expanding applicability to a range of difficult spinal procedures, the fiscal efficiency of spinal robotics will concomitantly increase.²⁶

8.3 Early Evolution of Spinal Robotics

The use of robotic assistance in spine surgery began in earnest in 2004 with the approval by the U.S. Food and Drug Administration of the Mazor™ SpineAssist system (Mazor Robotics, Caesarea, Israel). As is the case for the majority of spinal robotics systems to follow, the SpineAssist was largely designed to assist with the planning and implementation of a specific trajectory, whether it be for screws, needles, or other instruments. The device itself was mounted to the patient’s bony anatomy—either through a clamp attached to a spinous process, a bridge attached through three pins to the patient’s posterior superior iliac spine (PSIS) or a device mounted to the surgical bed. Planning took place through a reformatted CT scan of the operative area imported into a planning software suite. The C-arm was then calibrated and various intraoperative fluoroscopy images were taken and merged with the preoperative CT scan. Following this, a robotic arm directed the surgeon’s hand to the starting point and angle of the planned trajectory, and a sheath attached to the robotic arm served as a guide for the surgeon’s hand during instrument placement.²⁷

Several robotic devices intended for use in spine surgery are currently available, and each draws in one way or another on the functionality of SpineAssist; namely, to assist with planning and executing a trajectory with instruments and/or hardware. These include the Mazor™ Renaissance and Mazor™ X (Mazor Robotics, Caesarea, Israel), ExcelsiusGPS (Globus Medical, Inc., Audubon, PA), and ROSA (Medtech Surgical, Inc., New York, NY), among others (Fig. 8.1 and Fig. 8.2).

Fig. 8.1 The ExcelsiusGPS surgical system (Globus Medical, Inc., Audubon, PA). This system uses optical tracking and a patient-mounted reference array, obligating the need for the robot to attach directly to the patient or bed. (Image used with permission from Globus Medical.)

Only gold members can continue reading. Log In or Register to continue