13.1.1 Prevalence and Etiology

There has been a recent increase in spinal surgery with over one million spine procedures occurring annually. ¹ Of these patients, up to 40% will not meet their postoperative expectations of improvement and may suffer from persistent or recurrent back and/or radicular pain that is resistant to conservative measures. ² In addition, approximately 15% of patients that fail to improve will require revision surgery, with the majority attributable to postdiskectomy and postlaminectomy syndrome. ³ Pateder et al showed that in adult scoliosis patients with pseudarthrosis following multiple revisions, successful fusion was achieved in 90% of patients and was dependent on restoration of sagittal alignment, ⁴ yet overall, the evidence for improvement with surgical revision is variable. ^{2 ,​ 5 ,​ 6} Failure to improve from the initial surgery may be multifactorial. Underlying causes include preoperative factors such as poor patient selection and unrealistic postoperative expectations, intraoperative factors such as overly aggressive surgical decompression leading to spinal instability, and/or postoperative factors such as development of Flat Back Syndrome and pseudarthrosis. ^{2 ,​ 7}

13.1.2 Management

Success rates of revision surgery may decrease with each reoperation of spinal fusion with 50% success in the first reoperation, 30% in the second, and 15% in the third. ^{2 ,​ 5} Therefore, optimal medical management should first be considered, including efficient pain control and physical therapy; however, deteriorating neurological function or overt hardware malfunction causing instability may necessitate treatment. Patient evaluation should be performed to determine candidacy for less-invasive measures such as epidural injections or a spinal cord stimulator trial. If these strategies are not sufficient or fail, revision surgery should be considered as a treatment option.

13.2.1 Pedicle Screw

The majority of spine revisions, particularly in the thoracolumbar spine, are performed by a posterior or posterolateral approach and the most commonly utilized form of instrumentation is the pedicle screw. The pedicle screw instrumentation and constructs has the advantage of improved biomechanical stability due to its resistance to fatigue and increased pull-out strength. Because the pedicle screw is designed to span all three columns of the vertebrae, it provides overall support not offered by other fusion constructs (e.g., Harrington rods, wiring). ^{8 ,​ 9} Traditionally, pedicle screw insertion is performed freehand, guided by a meticulous knowledge of bony anatomy and appreciation for key landmarks exposed during surgery, such as the pars articularis, mammillary body, and the transverse process (e.g., Roy-Camille method). This method can carry a breach rate of 55% associated with a 1 to 8% risk of injury to surrounding structures such as nerve roots, adjacent viscera, and vasculature. ^{10 ,​ 11} This risk is amplified in revision cases where bony elements have been completely removed or modified and replaced with a fusion mass over time, making anatomical landmarks indiscernible. ¹⁰ In addition, surrounding scar tissue makes dura and nerve roots harder to delineate which escalates the risk of cerebrospinal fluid leak and of damage to underlying nerve roots. ^{8 ,​ 12}

13.2.2 Determining Pedicle Screw Breach Rate

Breaches may reflect a violation of instrumentation out of the bony confines. Most studies that evaluate the accuracy or breach rate of pedicle screws in the spine do so with the aid of grading systems that classify breaches based on several features observed on postoperative imaging. These features include the extent of screw perforation outside of the pedicle (mm), directionality of the screw breach (medial vs. lateral), as well as degree and type of associated neurological deficits, if present (Fig. 13‑1). Below are several grading systems for pedicle breaches used to determine accuracy of techniques for screw placement (Table 13‑1). Variations do exist among the breach grading systems. These differences should be appreciated as they may affect the ability to compare accuracy rates of screw placement methods across studies.

13.3.1 Pros and Cons of CT-Guided Computer-Assisted Navigation

Image-guided techniques for placement of posterior instrumentation in the spine ranges from plain serial radiography, C-arm fluoroscopy, to 2D Fluoro/3D CT-guided computer-assisted navigation (CAN). Furthermore, 3D navigation systems may use either preoperatively or intraoperatively acquired imaging for registration. Specifically, CT-guided CAN offers several advantages in the placement of posterior instrumentation (Table 13‑2). As mentioned, this technique allows the acquisition of multiplanar images that readily interface with the externally referenced navigation system. This eliminates the time-consuming and error-prone step of manual registration as the images are automatically registered and formatted on the workstation for use during navigation. Preoperative gantry settings can be adjusted to provide as low as reasonably achievable (ALARA) image acquisition resulting in lower doses of radiation exposure. ¹⁷ Planning of screw trajectory can be done in real time, which also allows for instantaneous adjustments of implants during the procedure. Further, the ability to acquire images intraoperatively while the patient is in final position using mobile CT technology reduces error introduced by positional changes that occur when images are acquired preoperatively with the patient positioned supine. ¹⁸ Use of CT-guided CAN may result in a significantly larger screw to pedicle ratio that enhances pull-out strength of implanted constructs and, indirectly, decreases the risk of future revision. ⁸ These advantages are not just limited to the thoracolumbar spine but have been proven in the placement of pedicle screws in the cervical spine as well. ¹⁹

Despite the advantages, several shortcomings have been suggested. These include longer operative times amplified by the learning curve necessary for surgeons and staff to become familiar with the navigation procedure, more exposure to radiation, and overall increased costs associated with purchase and upkeep of a navigation system. In addition, despite the numerous studies that have shown decreased breach rates with image guidance, these results have not translated into significantly improved clinical outcomes. Technical aspects of navigation have been implicated in hindering the accurate placement of pedicle screws as well. These include accidental shifts in position of the DRA after registration due to presence of abutting scar tissue or jolting from hand movements during space-occupying steps such as tapping or screw placement. ²⁰ Also, in patients with previous fusion implants, metal artifact can impede visualization of bony anatomy on imaging. ²¹ Yet, as the technology continues to evolve, evidence of these proposed shortcomings are becoming increasingly invalidated. ^{17 ,​ 20 ,​ 22 ,​ 23 ,​ 24 ,​ 25 ,​ 26 ,​ 27 ,​ 28 ,​ 29}