9 Outcomes in Navigated Spinal Surgery

10.1055/b-0039-172720

9 Outcomes in Navigated Spinal Surgery

Brian T. Sullivan, Alex A. Johnson, Nicholas Theodore, Jeff Jacobson, and A. Jay Khanna

Abstract:

This chapter reviews the major outcomes observed in modern spine surgery using navigated technologies. The aim of this chapter is to critically analyze current radiographic, clinical, surgical, patient-reported, and financial outcomes in computer-assisted navigated spine surgery compared with conventional fluoroscopic techniques. Another chapter will describe outcomes in robotic spine surgery. Navigation in spine surgery may be considered as an adjunct for optimizing surgical outcomes. However, a firm understanding of the anatomy and surgical skills needed for more conventional techniques remains important to the modern spine surgeon. Pedicle screw insertion with navigation has largely been shown to be more accurate and requires less time than conventional freehand techniques. Navigated spinal surgery decreases radiation exposure to the surgical team. Operative times in navigated spine surgery have shown noninferiority metrics when compared to conventional methods. Perioperative complications and patient-reported outcomes may require additional studies to better understand the complex interaction of navigation techniques and associated outcomes. Early comparative studies indicate less blood loss and the potential for decreased neurovascular intraoperative complications resulting from more accurate instrumentation placement. Some of the limitations and challenges seen in navigated spine surgery include high acquisition costs of the equipment, the learning curve for surgeon and healthcare team, and unique risks associated with the technology.

9.1 Introduction

Spine surgery requires meticulous dissection, a comprehensive understanding of spinal anatomy, and constant focus by the surgeon, especially when working near-critical neurovascular structures. Technological advances, including navigated and robotic spine surgery, were developed to reduce the chance for human error and improve three-dimensional (3D) visualization of anatomy in the setting of more minimally invasive approaches. Historically, spine surgeons relied on experience and open approaches with direct visualization of anatomical landmarks throughout the case to insert implants and used serial radiographic evaluation to confirm proper placement of spinal instrumentation. ¹ ^, ² ^, ³ This cumbersome process and high pedicle screw malposition rate ranging from 8 to 50% ⁴ ^, ⁵ ^, ⁶ ^, ⁷ led to the development of image-guided surgery, aiming to expedite the operation and improve the accuracy and precision of implant insertion. Over time, computer-assisted navigation and robotic-assisted techniques have become commonplace in spine surgery.

Computer-assisted navigation technology relies on computed tomography (CT) or fluoroscopic imaging in combination with specialized cameras and tracked surgical equipment that are processed to create a 3D visualization to guide the surgeon intraoperatively (Fig. 9‑1). Often, CT-based navigation requires preoperative registration and preparation when compared to fluoroscopy-based navigation (Fig. 9‑2). The theoretical advantages of computer-assisted navigation include real-time 3D assessment of anatomy, reduction in intraoperative fluoroscopic use and radiation, decreased operative time, and improved accuracy and precision of implant placement.

**Fig. 9.1** Operating room setup and workflow for a 79-year-old female with history of thoracolumbar scoliosis, multilevel lumbar stenosis, and L4–L5 spondylolisthesis who underwent a T10 to pelvis posterior spinal arthrodesis with instrumentation, L2 to L5 laminectomies, and L5–S1 transforaminal lumbar interbody fusion using intraoperative CT-based navigation (Brainlab navigation system) in partnership with orthopaedic surgery and neurosurgical colleagues at our institution. (a) The surgical team carefully places the patient in the prone position with appropriate supports and padding on the adjustable, image-guidance table. (b) The large operating room is shown, which supports the numerous personnel and equipment needed for appropriate workflow using navigated techniques. (c) The surgeon and his first assistant perform the initial exposure to mount the anatomic reference array and identify key anatomy in standard fashion. (d) The patient undergoes an intraoperative CT using the Airo Mobile scanner after the wound site and sterile field are draped with clear, plastic covering (to allow for visualization of the reference array during the scan). These images were used for intraoperative navigation and image guidance after accuracy and precision of image acquisition was confirmed on the Brainlab stereotactic workstation. (e) The surgeons use the intraoperative guidance for pedicle screw insertion with the stereotactic video monitor displayed in the upper field. (f) After placement of all of the pedicle screws, the surgeons begin the decompression portion of the procedure; in this case, the surgeons use an osteotome during the initial portion of the decompression. Note that an array is still attached on the S1 spinous process. A final CT scan was performed intraoperatively, and position of the instrumentation was noted to be excellent.

**Fig. 9.2** Steps required for accurate and precise image registration along with the planned trajectories of instrumentation using the Brainlab navigation system. (a) The coronal, axial, and sagittal views are shown to verify the registration accuracy based on comparison with known anatomical markers by the surgeon. (b) Each instrument has three attached reference points and is calibrated for use in image-guided surgery. This displays the calibration process for a navigated probe with reference markers to confirm accuracy. (c) Once the image registration and instrument calibration are confirmed, the planned trajectory for instrumentation is available in real time with multiplanar views via stereotactic guidance.

Robotic spine surgery, an extension of navigation techniques, is a broad term that encompasses any operation that partially or completely utilizes robotic devices or artificial intelligence. Robotic surgery offers the advantage of improved precision of movement beyond the capacity of human physiology. ⁸ ^, ⁹ Robotic surgery is introduced fully in Chapter 2 and explored throughout the remainder of this book.

Although the technology available in the modern era of navigated spine surgery is exciting and shows promising clinical and radiographic potential, it is of paramount importance for spine surgeons to understand the efficacy and limitations of this technology in terms of clinical outcomes compared to traditional unguided methods. This is particularly true in an age when cost and quality of care are increasingly scrutinized. The aim of this chapter is to critically analyze current radiographic, clinical, surgical, patient-reported, and financial outcomes in computer-assisted navigated spine surgery compared with conventional techniques. The merits and downfalls of specific computer-assisted navigation platforms are beyond the scope of this discussion and have been discussed elsewhere. ¹⁰

9.2 Discussion

The following discussion explores some of the key outcomes in navigated spine surgery, including, but not limited to, pedicle screw accuracy rates, radiation exposure to the patient and operating room personnel, operative time, perioperative complications, and patient-reported outcomes (PRO). The limitations of navigated spine surgery, such as initial acquisition costs, learning curve, and unique risks, are reviewed thereafter.

9.2.1 Pedicle Screw Accuracy Rates

Pedicle screws are now a mainstay in spine surgery due to the structural integrity offered with three-column fixation, rotational control, and increased pullout strength. Correctly positioned pedicle screws avoid high-risk intraoperative complications including vascular injury, devastating neurological sequelae, and the need for revision surgery associated with other implants. Evaluation of pedicle screw placement is one of the most studied aspects in the navigated spine surgery literature.

In a systematic review of 30 studies, Mason et al found 3D fluoroscopic navigation-guided (96%) modalities to be more accurate than two-dimensional (2D) navigation-guided (84%) and conventional fluoroscopy counterparts (68%). ⁶ In a meta-analysis of 10 studies comparing 3D CT-based navigation and 2D- and 3D-fluoroscopy-based navigation techniques, Du et al found significant improvements in the screw placement accuracy with the 3D fluoroscopy navigated group when compared to the CT navigation group (relative risk [RR] 95%, CI: 0.4–0.9, p = 0.01) and 2D navigation group (RR: 95%, CI: 0.2–0.6, p < 0.01). ¹¹

Several additional meta-analyses using mostly individual cohort studies along with prospective and randomized controlled trials to evaluate pedicle screw accuracy rates found improved accuracy rates with navigation techniques over nonnavigated counterparts (Table 9‑1). However, not all studies agree with the result, as one study found no difference in 3D CT-based navigation pedicle screw accuracy (96%) compared with freehand placement (96%) in a retrospective review of 49 subjects with idiopathic scoliosis who underwent placement of a total of 835 pedicle screws. ¹²

**Table 9.1** Recent meta-analysis and systematic review studies investigating pedicle screw position accuracy rates in spine surgery using conventional freehand surgery (fluoroscopy) versus navigated techniques
Study, author (year published)	Conventional (%)	Navigated 2D or 3D fluoroscopy (%)	Navigated CT (%)
Kosmopoulos and Schizas (2007) ¹³	90.3	NR	95.2
Tian and Xu (2009) ¹⁴	NR	85.5	90.8
Verma et al (2010) ¹⁵	84.7	NR	93.3
Shin et al (2012) ⁷	85.0	NR	94.0
Mason et al (2014) ⁶	68.1	2D: 84.3 3D: 95.5	NR
Bourgeois et al (2015) ⁴	86.9	NR	99.7
Meng et al (2016) ¹⁶	84.9	NR	94.6
Abbreviation: NR, not reported.

Despite the overall increase in accuracy when compared to freehand techniques, Uehara et al ¹⁷ showed in a review of 359 subjects that CT-based navigation may be more accurate in the lumbar spine than in the cervical spine in which they reported significantly higher pedicle screw breach rates. Similarly, in the thoracic spine region, Kosmopoulos and Schizas ¹³ found no differences in pedicle screw accuracy rates in a meta-analysis and subgroup analysis comparing navigation with nonnavigated techniques, despite an overall improvement in the navigated group. These findings suggest that navigation techniques remain as a surgical adjunct, and an exquisite understanding of the anatomy is paramount for successful spine surgery. Confounders in comparing accuracy rates across studies may be limited by the surgeon’s experience level, open versus minimally invasive approaches, patients’ body habitus among other clinical variables, and the criteria or diagnostic tools used to identify pedicle screws in malposition. ¹⁸

Unique Populations

Congenital scoliosis: In this group of pediatric disorders which are frequently associated with challenging anatomy, one study reviewing pedicle screw placement in 14 consecutive children with congenital spine deformity found that CT-based navigated screw placement resulted in a 99.3% accuracy rate and offered the benefits of identifying missing or dysmorphic pedicles intraoperatively. ¹⁹
Tumor: Image guidance has largely become standard of care for cranial oncology surgeries. As oncology surgeries often require complete resection of tumor across multiple tissue planes, intraoperative image guidance has been shown to improve accuracy of resection in spinopelvic regions. ²⁰ ^, ²¹ ^, ²² D’Andrea and colleagues ²³ reported on a unique technique which combines preoperative MRI with intraoperative CT-based navigation to optimize visualization of tumor burden and surrounding neurovascular structures in the spine. This co-registered navigation technique effectively permitted the complete resection of tumor burden at 1-year follow-up but was limited by the small case series size (n = 4) and heterogeneous group of pathology.
Minimally invasive spine surgery (MISS): In MISS, the surgeon has limited direct visualization of the relevant anatomy and landmarks and must rely more on visual and radiographic adjuncts. Three-dimensional spine navigation uniquely aids the surgeon by providing visualization with limited exposure. The surgeon has real-time feedback and unlimited degrees of freedom with his or her tools, and several studies indicate improved instrumentation accuracy rates in the setting of MISS with 3D navigation guidance. ⁴ ^, ²⁴ ^, ²⁵ In addition, the benefit in terms of radiation dose reduction is greatest in minimally invasive spine surgical procedures given the greater amounts of radiation that are typically required with conventional fluoroscopic guidance.