Introduction of Technology: Adult Spinal Deformity
Background of Adult Spinal Deformity
Adult spinal deformity (ASD) is defined as an abnormal curvature or abnormal alignment of the vertebral column. ASD has been demonstrated to limit a patient’s quality of life significantly, and patients with both cervical and thoracolumbar deformity have predictable declines in health-related quality of life (HRQOL). In the most severe cases, patients with lumbar scoliosis and sagittal malalignment have quality of life measures equal to patients who have lost the ability to use their upper or lower extremities. The definition of scoliosis is a Cobb measurement greater than 10 degrees; however, in de novo adult degenerative scoliosis, coronal deformity is considered when Cobb measurements exceed 20 to 30 degrees. The Scoliosis Research Society (SRS) Schwab classification for spinal deformity classifies major coronal deformity as thoracic, thoracolumbar, or lumbar scoliosis greater than 30 degrees.
A globally aligned spine allows for the least amount of energy expenditure necessary to maintain upright posture, bipedal gait, and horizontal gaze. In the globally aligned spine, a C7 plumb line should fall between the posterior superior corner of the sacrum and the center of femoral heads, and the central sacral vertical line (CSVL) should transect the C7 spinous process. While coronal imbalance greater than 4 cm has been associated with increased pain and decreased function, sagittal imbalance correlates more closely with HRQOL and is the primary focus of adult deformity correction.
The importance of sagittal alignment was introduced with the conception of pelvic parameters and became the cornerstone for spinal deformity correction, as it was shown to correlate with the HRQOL measures. The body attempts to compensate for sagittal imbalance through a series of mechanisms including reduction of thoracic kyphosis, retrolisthesis of lumbar vertebra, pelvic retroversion, and knee flexion. It is important to identify these clinical and radiographic compensatory mechanisms as they can mask underlying spinal deformity. The most common spinopelvic parameters used to describe global sagittal alignment are pelvic incidence (PI), sacral slope (SS), pelvic tilt (PT), lumbar lordosis (LL), and sagittal vertical axis (SVA). Similarly, the C2–C7 SVA and the chin-brow vertical angle (CBVA) are used to assess the degree of cervical spinal deformity. In order to attain the greatest improvement in HRQOL, the following postoperative parameters should be achieved: PT < 25 degrees, PI-LL mismatch + 11 degrees, and SVA < 5 cm. Historically, the treatment of spinal deformity had been supportive and aimed at delaying the progressive decline in a patient’s overall health. However, currently, the treatment is corrective and aimed at improving the patient’s quality of life. Significant evidence has been published demonstrating a clinically significant improvement in the HRQOL measure following the surgical correction of ASD.
The potential to improve the quality of life of patients with spinal deformity significantly through surgery are well appreciated. However, it is biased to discuss the potential benefits of surgical intervention without discussing the high complication rate associated with performing these surgeries. In a two-year prospective multicenter study, the complication rate following ASD correction was approximately 70% and the revision surgery rate was approximately 28%. The most frequent complications following ASD surgery include rod breakage, proximal junctional kyphosis (PJK), postoperative anemia, surgical site infection, and neurologic injury. Currently, predictive analytic models are being developed to identify patients with spinal deformity that would benefit from surgical correction and are at low risk for complication. The use of modern technology with machine-based learning to predict success rates following ASD surgery is a perfect example of how modern technology is improving the field of spine surgery.
Spinal Navigation and Robotics in Spinal Deformity
Since 1911, when the first spinal fusion was performed, technology has been the fuel propelling advances in spine surgery. Engineers and like-minded inventors are perpetually advancing the field, and with developments in imaging modalities, interbody instrumentation, minimally invasive surgical (MIS) techniques, biologics, navigation software, and robotics, the face of spine surgery is constantly in flux. Despite the many avenues of technologic growth in spine, technologic advances in the last 30 years have been dominated by improving imaging modalities, redefining stereotactic image guidance, and introducing robotic assisted surgery. The Oxford dictionary defines “navigation” as the process or activity of ascertaining one’s position and planning or following a defined route. The concept of navigation is as much at the foundation of spine surgery as it is the basis for nautical or aerospace travel. At its core, spine surgery relies on visible anatomic landmarks and the mind’s eye to recreate a three-dimensional (3D) mental representation of the spine to navigate surgery safely. Intraoperative imaging technologies have expanded the ability of the spinal surgeon to visualize the 3D anatomy of the spine, relying less on presumptions based on visible landmarks. Spinal navigation expands on intraoperative imaging, utilizing stereotactic technology to allow the surgeon to visualize their instrumentation in 3D radiographic images of the patient’s spine. Robotic shared control devices represent the current pinnacle in navigation technology. With robotic technology, the surgeon plans the pedicle screw trajectories on pre- or intraoperative imaging; these are subsequently produced by an automated robotic arm.
The ultimate goal of using computer assisted spinal navigation and robotics in ASD is to improve the precision and accuracy of the spine surgeon, which will lead to fewer complications and improved surgical outcomes. The theoretical advantages of navigation and robotics are (1) improved accuracy of pedicle screw placement, (2) improved ability to perform minimally invasive spine surgery, (3) decreased radiation exposure to the spine surgeon, and (4) avoiding surgeon fatigue and human error. The remainder of this chapter will focus on the application of navigation and robotics for use in ASD.
Accuracy of Pedicle Screw Placement
The accuracy of pedicle screw placement has been the most investigated topic for both spinal navigation and robotic assisted surgery. In deformity surgery, the placement of accurate pedicle screws becomes increasingly more difficult because of the loss of anatomic landmarks in revision settings, rotatory scoliosis requiring unique pedicle screw trajectories, and the presence of dysplastic pedicles ( Fig. 4.1 ). Spinal navigation and robotics are tools that can facilitate accurate screw insertion in the aforementioned difficult clinical scenarios, and the use of CT spinal navigation has been shown to have equivalent pedicle screw insertion accuracy in primary and revision clinical cases.
A number of studies have been performed comparing the accuracy of pedicle screw placement between free-hand techniques, fluoroscopic techniques, navigated techniques, and robotic techniques. A large meta-analysis compared pedicle screws placed using either conventional fluoroscopy, 2D fluoroscopic navigation, or 3D fluoroscopic navigation, and found accuracy rates of 68.1%, 84.3%, and 95.5%, respectively. In a randomized controlled trial (RCT) that compared 100 consecutive patients with pedicle screws placed using either conventional techniques or navigation, the pedicle breach rate was 13.4% in the conventional group and 4.6% in the navigated group. A prospective RCT of 1116 thoracolumbar pedicle screws demonstrated that robotic assisted pedicle screw placement was associated with better accuracy, a lower rate of violating the proximal facet capsule, and a lower rate of medial pedicle breach compared to the free-hand technique. Similar results of improved accuracy with spinal navigation have been reported in the deformity setting. An RCT comparing navigated and nonnavigated pedicle screw accuracy in patients with spinal deformity demonstrated a pedicle breach rate of 2% with navigation and 23% without navigation. In a review of patients with neuromuscular scoliosis and dysplastic apical pedicles, the accurate placement of apical pedicle screws was achieved in 79% of pedicles utilizing navigation and 67% of pedicles without navigation. A systematic review compared navigated and non-navigated pedicle screw insertion in scoliosis patients and found a higher pedicle screw perforation rate without the use of navigation. However, no difference in revision treatment rates for screw malposition was seen between the groups.
In addition to thoracolumbar pedicle screws, navigation has been shown to be effective for the instrumentation of cervical pedicle screws and lumbopelvic fixation. A number of technical guides and case series have demonstrated the usefulness of CT-guided navigation for the placement of S2-alar-iliac screws. Cervical pedicle screws are technically demanding with a lateral breach threatening the vertebral artery and a medial breach placing the spinal cord at risk. The free-hand technique for subaxial cervical pedicle screws hm 14.3% to 29.1%. Cervical pedicle screw placement using 3D navigation has an accuracy of 89.7%. In the upper cervical spine, CT navigation has been shown to be an effective tool for the placement of C1 and C2 instrumentation. However, the single head-to-head study comparing free-hand to navigated techniques for C2 pars screw placement found higher accuracy with the free-hand technique.
Navigated MIS Applications in Deformity
In ASD there is a growing trend toward utilization of MIS techniques for deformity correction. MIS techniques decrease the intraoperative blood loss, overall complication rate, and length of hospitalization. However, it is important to select suitable patients, as under-restoration of deformity parameters compared to conventional open surgery is frequently reported. A classification system has been developed to aid in the selection of appropriate patients whose spinal deformity can be corrected successfully with MIS techniques. In general, fixed sagittal plane deformities should be considered cautiously for purely MIS techniques, and these deformities are best addressed with open or hybrid techniques. The most common MIS technique for deformity correction involves a two-position surgery with a lateral lumbar interbody fusion (LLIF) and percutaneous pedicle screws. A hybrid technique has been described with an LLIF followed by an open posterior fusion incorporating Smith Peterson osteotomies, allowing for greater sagittal correction. A relatively recent MIS technique, anterior column realignment (ACR), allows for greater restoration of LL by releasing the anterior longitudinal ligament (ALL). A promising adoption of navigation and robotic assisted surgery allows for single-position surgery placing both the LLIF cage and percutaneous pedicle screw from the lateral decubitus position. Single-position surgery eliminates the need to stage the surgery, change the patient’s position intraoperatively, or potentially require two intraoperative CT scans if navigating both the interbody cage and percutaneous pedicle screws.
Visualization for the accurate placement of lateral interbody cages and percutaneous pedicle screws has traditionally, and successfully, been accomplished with conventional fluoroscopy. However, it has been demonstrated clearly that fluoroscopy in MIS surgery exposes the surgeon to high levels of ionizing radiation, and that 3D navigation greatly reduces the radiation exposure to the surgeon. Initial studies investigating CT navigated LLIF demonstrated that navigation allows for successful and reproducible placement of the interbody cage. With the use of CT navigation, the instruments required to prepare the disk space, trial different size cages, and place the actual cage can all be navigated in real time.
Percutaneous pedicle screws can be inserted using conventional fluoroscopic imaging, 2D or 3D navigation, or robotic assisted navigation. The largest study to date reported an accuracy rate of 97% with robotic assisted placement of percutaneous pedicle screws. One study compared conventional percutaneous screw placement, 2D navigated percutaneous screw placement, and 3D navigated percutaneous screw placement and reported screw malposition rates of 5.16%, 7.29%, and 1.23%, respectively. In addition to the reduced radiation exposure with navigated percutaneous pedicle screw insertion, navigated and robotic assisted placement of screws minimize the risk of violating the proximal facet joint capsule. Damaging the proximal facet capsule can lead to increased rates of adjacent segment deterioration and PJK. A cadaver study demonstrated violation of the proximal facet joint capsule in 58% of those percutaneous screws placed with conventional fluoroscopy. LLIF and percutaneous pedicle screw insertion can be performed safely and effectively with conventional fluoroscopic imaging or 3D navigated technologies. However, the utilization of navigation significantly lowers the radiation doses to both the operating surgeon and the patient, and preservation of the proximal facet joint capsule is more likely to occur when using robotics and navigation.
Unique Applications of Spinal Navigation and Robotics in Deformity
At present, robots are assistive tools for the accurate placement of percutaneous and open pedicle screws. Robotic technology can aid the deformity surgeon by allowing the accurate placement of pedicle screws in the challenging setting of rotated and dysplastic pedicles. While the current depth of utilization with robotics is pedicle screw insertion, the applications of spinal navigation for deformity surgery are much broader. In addition to pedicle screw insertion, spinal navigation has been used for the planning of cervicothoracic pedicle subtraction osteotomies (PSOs), for determining the resection planes for lumbar PSO, and for 3D visualization of the patient’s head alignment in space. Case reports in the literature have described how spinal navigation can be used to plan resection planes and navigate osteotomes for the accurate execution of lumbar PSOs. In the revision setting, previous fusion masses mask the lumbar anatomy, exponentially increasing the difficulty of performing three-column osteotomies. Navigation can assist in safely performing three-column osteotomies through previous fusion beds. In addition to lumbar osteotomies, a case report demonstrated the utility of spinal navigation for cervical spine PSOs. Multiple studies have reported the utility of spinal navigation with tumor resection; similarly, within the deformity literature, spinal navigation has been shown to assist with vertebrectomies of hemivertebra, resulting in congenital scoliosis. As demonstrated in our case report, intraoperative CT can be utilized to create 3D representations of the patient’s head in space. This application is useful when correcting cervical deformity or performing occipitocervical fusions. In our case, we were able to measure the CBVA successfully intraoperatively to ensure adequate restoration of horizontal gaze following deformity correction. It is of the utmost importance when fusing the patient’s head in a fixed position or correcting cervical deformity to position the head such that the patient can visualize the world with the greatest ease. It has been reported that the CBVA following deformity correction should be between 10 and 20 degrees to provide the greatest functional outcome. The ability to effectively measure the CBVA intraoperatively on a 3D image of the patient’s head-neck relationship ensures the appropriate head position prior to spinal fusion.
Limitations of Spinal Navigation
Despite stereotactic navigation being invented in 1908 and spinal navigation coming onto the market in the 1990s, spinal navigation for spinal deformity remains in its infancy. Many surgeons are resistant to using the technology, and not without right cause. The navigation is based on referencing, and movement of the patient or referencing frame following registration will result in inaccuracy of the location of the navigated instrument in relation to the patient’s anatomy. This limitation of the technology needs to be understood by the operating surgeon, and it remains imperative to assess the accuracy of the navigation frequently throughout a surgical case. The use of intraoperative navigation or robotic assisted surgery requires additional equipment and complexity for room setup and workflow. In order to not significantly delay operative times, the OR staff must become proficient in the workflow of using these technologies. In the current era of rapidly evolving technology, it is exciting to see where navigation and robotics will lead in the field of spinal deformity. With advances in technology, the goal remains to minimize complications and maximize patient outcomes following deformity correction.
A 66-year-old female presented with the complaint of significant deformity of her neck and progressive symptoms of worsening balance and poor hand dexterity. Her previous surgical history was significant for a posterior cervical laminectomy for intramedullary spinal cord ependymoma in 1970. On examination she demonstrated intact motor and sensory function but exaggerated deep tendon reflexes and gait instability. She was unable to look upright.
Preoperative cervical radiographs demonstrated fixed kyphosis measuring 91.5 degrees ( Fig. 4.2 ). A subsequent CT scan of the cervical spine confirmed fused facet joints from C1 to C5. A radiographic diagnosis of post-laminectomy kyphosis was confirmed. The patient’s SVA was 1.7 cm, the C2–C7 SVA equaled 3.5 cm, and the CBVA was 22 degrees ( Fig. 4.3 ).