Kristen Radcliff, MD,1 Harvey Smith, MD,2 Bobby Kalantar, MD,3 Robert Isaacs, MD,4 Barrett Woods, MD,1 Alexander R. Vaccaro, MD, PhD, MBA,1 James Brannon, MD5
1Department of Orthopedic Surgery, Thomas Jefferson University, Philadelphia, PA 2Department of Orthopedic Surgery, University of Pennsylvania, Philadelphia, PA, 3Department of Orthopedic Surgery, Georgetown University, Washington, DC 4Department of Neurological Surgery, Duke University, Durham, NC 5Orthopedic Sciences, Inc, Seal Beach, CA
Perforations of the pedicle wall during cannulation can occur with experienced surgeons. Direct endoscopic visualization has not been used to inspect pedicles previously due to bone bleeding obscuring the camera visualization. The hypothesis of this study was that endoscopic visualization of pedicle wall integrity was technically feasible and would enable identification of clinically significant pedicle breaches.
A live porcine model was used. Eight lumbar pedicles were cannulated. Clinically significant breaches were created. An endoscope was introduced and was used to inspect the pedicles.
All lumbar pedicles were endoscopically visible at a systolic pressure of 100 mm Hg. Clinically relevant anatomic structures and iatrogenic pathology, such as medial, lateral, and anterior breaches, were identified. There were no untoward events resulting from endoscopic inspection of the pedicle endosteal canal.
Endoscopic inspection of lumbar pedicles was safe and effective. The findings on endoscopic inspection corresponded with the ball-tip probe palpation techniques. Additional techniques, such as selection between two tracts, was possible with the endoscopic technique.
Pedicle screws have revolutionized the treatment of spinal disorders. With screws, surgeons are able to immobilize and manipulate the spine in three dimensions. Pedicle screw instrumentation is the standard of care in the surgical management of degenerative scoliosis, degenerative spondylolisthesis, and trauma. To place pedicle screws, surgeons utilize handheld instruments to displace soft, cancellous bone within the pedicle whilst simultaneously preserving the external hard bony wall of cortical bone. The surgeon uses carefully defined anatomical landmarks and tactile feedback to create a screw tract through the cancellous bone of the pedicle. Once positioned, the screw position is often checked for encroachment upon spinal nerves by measuring the electrical conduction through triggered electromyography. Intact cortical walls create significant electrical impedance.1 Correctly placed pedicle screws will require higher amperage to detect electrical activity in adjacent spinal nerves.1 Other techniques, such as fluoroscopy2 or intraoperative CT scan,3,4 are also used by some surgeons to guide or check screw placement.
Perforations of the outer cortical pedicle wall can occur,5 using manual free-hand techniques,6 fluoroscopy,2,7 and even intraoperative CT.3 Perforation rates as high as 30% have been reported in some studies.8 Pedicle screw malposition can result in unanticipated readmission and reoperation,9-11 dural laceration,12 nerve injury,13,14-17 pedicle fractures,17,18 and vascular injury.19,20
Direct visualization of the cannulated bony channel would provide valuable information to confirm accurate trajectory of the tract and documentation for the medical record confirming the absence of a cortical wall breach. Furthermore, if a cortical wall breach is observed endoscopically, a handheld ball-tip probe can be used to palpate the direct area of concern versus complete reliance of proprioception and blind palpation. Endoscopy has revolutionized visibility of other hard-to-see locations within the body such as the knee, shoulder, and abdomen. Endoscopic visualization is superior to traditional open surgery that relies on line of sight vision in many cases. However, endoscopy has not been widely used in the placement of spinal instrumentation due to resident bleeding within the pedicle obscuring visualization, the small diameter in which visualization must occur, and the inability to exploit the unique property difference between a Newtonian Fluid (water) and non-Newtonian Fluid (blood).21,22,23,24
Recently, an endoscopic instrument (The Beacon, Orthopedic Sciences Inc, Seal Beach, CA) was developed to facilitate endoscopic inspection of the internal pedicle channel. The hypothesis of this study was that endoscopic visualization of pedicle wall integrity is technically feasible and will enable identification of clinically significant pedicle breaches.
One skeletally mature female pig (approx. 180 lb) underwent posterior lumbar exposure and pedicle cannulation followed by endoscopic verification of pedicle wall integrity. The investigation was performed using an approved IACUC protocol at an accredited facility. The investigators were two orthopedic surgeons with familiarity in intraoosseous endoscopy and spine surgery.
General anesthesia was induced. The animal was anesthetized according to veteranian’s protocol. An endotracheal tube was attached to an anesthesia machine. Replacement fluids (0.9% NaCl) were administered via the intravenous catheter. An area on the abdomen was shaved to accommodate an electrode return patch. The animals were placed prone on the surgical table. The area around the lumbar spine was shaved, prepped, and draped in preparation for surgery. Normal systolic and diastolic blood pressure was maintained throughout the procedure (mean arterial pressure 70). An arterial line monitor was used to monitor blood pressure.
Midline, posterior open dissection was performed through the skin and subcutaneous tissues on the posterior lumbar spine at the level of the iliac crests. Subperiosteal exposure of the spinous processes, laminae, facets, and transverse processes was performed. Meticulous hemostasis was obtained.
The pedicles were then cannulated in the usual fashion. The upslope of the transverse process onto the superior articular process was identified. The intersection of a horizontal line through the midpoint of the transverse process and the lateral border of the facet was considered the starting point of the pedicle screws. The starting point was decorticated with a rongeur. The pedicle was entered with a curved, tapered gearshift with pronosupination in the usual fashion. When necessary, a radiograph was taken to confirm the trajectory and spinal level. The pedicles were cannulated 35mm according to the depth on the gearshift. The pedicles were entered 35mm according to the depth on the gearshift. Eight holes were made in the spine with an attempt to establish correct pedicle canals holes as well as to breach pedicle canals.
Following cannulation, the inner aspect of the pedicles was palpated with a ball-tipped probe as carefully as possible. A consensus was reached between the investigators about whether the pedicle was intact or breached. If a breach was suspected, the direction was also reported.
The endoscopic instrument outer trochar was connected to normal saline irrigation in 3 Liter bags. The endoscopic instrument uses commonly available endoscopy monitors available in most hospitals. No epinephrine was present in the normal saline bags. The endoscopic instrument (with inner stylette and outer trochar together) was then placed into the pedicle tract. Saline was allowed to flow at gravity pressure. No specialized pumps or pressure bags were used. The inner stylette, which has a diameter of 3.2 mm, was removed. A 3.0 mm endoscope was then introduced into each pedicle. The pedicle wall was inspected with a 0 degree and a 30 degree endoscope.
Breaches were deliberately made in the medial wall, anterior vertebral body, and lateral muscle tissue. The breaches were confirmed by palpation with a ball-tip probe.
The endoscopic instrument was used to inspect the pedicle walls and confirm the breach locations.
Eight pedicles were cannulated in total in L6, L5, L4, and L3. All (8/8) of the pedicles were successfully visualized with the endoscopic instrument. The internal blood within the pedicles was cleared immediately once the endoscopic instrument was seated within the pedicle and irrigation was commenced. There were no pedicles (0/8) that were not able to be studied due to bleeding. Complete examination of the interior of each pedicle from posterior to anterior could be accomplished within one minute.
Using both the 0 degree scope, the medial, lateral, superior, and inferior walls and the floor were visualized on all pedicles (Figure 1, Video 1). The visualized pedicle wall integrity corresponded in all cases (8/8 pedicles) to the investigator’s assessment with ball-tip probe. The thirty-degree scope was used in four pedicles. The 0-degree scope was found to be sufficient to search for breaches in all cases. The major difference during endoscopic visualization between a 0-degree scope and a 30-degree scope is that a more direct view of the endosteal surface is achieved with a 30-degree scope. This direct wall inspection with the 30 degree scope may be helpful when a breach is identified and a more detailed inspection of the breach is required.
Breaches were deliberately made using the curved gearshift in the pedicles medially (n=2), laterally (n=2), and anteriorly (n=2). Perforations of the pedicular walls were easily identified with the endoscope in all six cases. Medially, the exposed dura and epidural fat could be visualized (Figure 2, Video 2A, Video 2B). Perforation of the anterior vertebral body was also easily visualized (Figure 3, Video 3) showing the anterior longitudinal ligament. Lateral perforation was also visualized, showing paraspinous muscles (Figure 4, Video 4a, Video 4b).
The endoscopic instrument was used in combination with a ball-tip probe to visualize the defect into which the ball-tip probe was subsiding (Figure 5, Video 5). A guidewire was introduced into a pedicle and the endoscopic instrument was used to document the position of the guidewire (Figure 6, Video 6). One pedicle was cannulated twice from the same starting point to create two trajectories. The endoscopic instrument was used to identify and explore dual paths in a lumbar pedicle (Figure 7, Video 7).
There was no significant extravasation of fluid into the spinal canal. There was no thecal sac compression visualized or nerve root compression visualized due to irrigation fluid. Once the irrigation was ceased, there was no retrograde flow of irrigation fluid from the extraosseus structures into the pedicles. There was no significant soft tissue swelling on breached cases. In two instances, irrigation was deliberately stopped. When irrigation was stopped, blood flow from the pedicle walls resumed immediately. (Figure 8, Video 8A, Video 8B). The blood was then quickly cleared with a resumption of irrigation (Figure 8, Video 8A, Video 8B).
Safe placement of spinal instrumentation is of paramount importance in overall successful spinal surgery. These results indicate that low pressure endoscopic inspection is technically feasible in a live animal model. Endoscopy provided valuable information about anatomical defects and guidewire placement in an efficient manner. There were no complications directly identified from the endoscopic technique. There were no specialized anesthetic requirements from the endoscopic technique.
Current techniques to verify pedicle screw placement are not universally accurate or effective. Manual pedicle palpation with a ball-tip probe has low accuracy25,26 and the potential for inadvertent neurological injury with the probe.27 Even stereotactic image guidance cannot entirely prevent pedicle perforations.28 Electrical neuromonitoring adds cost and time to surgical procedures. In some circumstances, neuromonitoring can fail to identify a pedicle screw breach.29,30 False positive alerts can also occur with neuromontoring, requiring additional operative time and steps. A recent systematic review concluded “There is no evidence to date that IOM can prevent injury to the nerve roots. Unfortunately, once a nerve root injury has taken place, changing the direction of the screw does not alter the outcome.”1 Other pressure and electrical conduction based techniques to verify pedicle accuracy, such as specialized piezoelectric piercer probes, can also lead to misplaced pedicle screws31 and also result in false-negative errors.32,33 Other, emerging techniques such as intraosseous ultrasound,34,35,36-39 robotic guidance, or near-infrared spectroscopy, require substantial equipment and are not without false negative errors.40,41
To our knowledge, our study is the first to report successful endoscopic visualization of the internal aspect of a lumbar pedicle with low flow irrigation. Endoscopy has been extensively used in other areas of surgery to minimize exposure. Current endoscopic spine techniques include foraminal discectomy or decompression techniques, but endoscopic instrumentation and fusion has not been widely performed due to technical challenges. Most previous descriptions of endoscopic pedicle screw instrumentation have focused on endoscopic soft tissue dissection, endoscopic identification of screw starting points, and endoscopic placement of rods.21,22,23 In the aforementioned studies, there is no description of inserting the endoscope into the pedicles to inspect the wall integrity.21,22,23
There are a few studies that describe visualization of the intraosseous anatomy of the lumbar pedicles with high flow irrigation.24,42,43 However, in contrast to the current technique, previous authors utilized high-pressure irrigation. In a high pressure irrigation environment, the endoscope is continually flushing away active bleeding in an open system. In our study, the unique design of the Beacon seals the pedicle and creates a closed system. In such a closed system, bleeding will not occur into the pedicle tract because the fluid cavity has been effectively sealed. Thus, only low flow irrigation is necessary to remove the static, motionless blood. The Beacon exploits the differing fluid properties between Newtonian (saline) and non-Newtonian (blood) fluids, thereby producing crystal clear images so that relevant anatomy can be viewed and documented. Figure 8 illustrates laminar streaks of blood (non-Newtonian fluid) flowing through the saline medium. Such laminar flow requires that the saline (a Newtonian fluid) within the pedicle is otherwise motionless. Thus the endoscopic instrument seals the pedicle channel, prevents fluid from escaping from the pedicle, and creates a closed system. If the saline were flowing in a high pressure environment, it would disrupt the laminations and thus turn the image cloudy. The current endoscopic instrument, through the attached fluid column, equilibrates the systolic pressure with only a minimal volume of saline in an intact pedicle. Therefore, minimal or low pressure irrigation is all that is necessary for successful instrumentation. High flow irrigation could create complications such as edema, neural element compression, or compartment syndrome. High-pressure lumbar irrigation has been demonstrated to increase cervical epidural pressure and possibly lead to intracranial hypertension.44
In our study, clinically relevant breaches were identified. These breaches were confirmed by manual palpation under direct endoscopic visualization. In this manner, even the palpation with a ball-tip probe under direct visualization was more controlled and perhaps less likely to cause inadvertent neurological injury. There were no complications from the usage of the endoscopic technique. Furthermore, the time per pedicle was approximately one minute. Once the endoscopic instrument was seated within the pedicle, the entire intraosseous space was sealed. Therefore only a minimal pressure was needed to cease blood flow and in fact to create retrograde flow. The animal’s blood pressure was normal. No abnormal hemodynamic conditions (such as severe hypotension) were necessary such as severe hypotension to reduce bony bleeding.
Limitations of the current study include the use of a porcine model. However, porcine models have been used in spine surgery previously for feasibility studies.41,45-53 Two authors have a direct conflict of interest. Additional studies by other, non-conflicted, investigators are needed. Other limitations include the small number of pedicles tested. In contrast to previous studies, threads were not tapped into the pedicles in this study24 because many modern pedicle screw systems are now deemed “self-tapping.” Therefore, the experimental conditions were designed to simulate the current surgical technique. Another limitation is that actual pedicle screws were not placed in this study. We acknowledge that there is potential that pedicles could be correctly cannulated but that errant screws could be placed due to misdirection after cannulation. However, the purpose of this study was to enhance the ability to identify a correctly cannulated tract. Despite the assistive techniques currently available, pedicle screw revision is one of the leading causes of reoperation after spine surgery, particularly in the hyperacute postoperative period.10,11
The main advantage of using an endoscopic technique for lumbar pedicle trajectories is the direct visualization of pedicle wall integrity. This feature lends itself readily to photographic and video documentation and education of trainees. In a recent cadaveric study, a pedicle breach rate of 51% was reported with resident physicians attempting to cannulate thoracic pedicles.54 The endoscopic instrument does not require any special setup, placement of needles or electrodes into the patients, alternative anesthetic techniques, monitoring equipment, special training, a change in technique, or capital expenditures. The endoscopic instrument allows the surgeon to visually identify cortical breaches before compression and neurological injury by screws occur. The endoscopic instrument does not employ any ionizing radiation. An additional advantage is that the endoscopic instrument continuously irrigates the pedicles, with an antibiotic solution of the surgeon’s choice, and may thereby reduce the risk of surgical site infection. Other indirect techniques, such as electrical stimulation, have been associated with false positive and false negative errors. Electrical stimulation, the most widely used method to check pedicle screw placement, also requires specialized anesthetic technique involving the absence of chemical paralysis. The endoscopic technique is versatile and similar to an open technique, which allows surgeons a smoother learning curve. Importantly, if a treating surgeon continues to observe breaches, with the added benefit of the location of the breach, he/she could then make real time adjustments to the technique used to cannulate the pedicle as a way of improving patient safety. There is no ionizing radiation in the endoscopic technique. Further developments in endoscopic imaging and fusion techniques will help refine this procedure. Currently, the screw system does not permit full endoscopic placement of longitudinal connecting rods. Additionally, the technique is diagnostic only. The system does not currently incorporate endoscopic “drilling” features to cannulate a pedicle. Ultimately, we believe that endoscopically assisted posterolateral lumbar instrumentation will reduce perioperative complications, costs, and the risk of return to the operating room.
Video 1A: Endoscopic pedicle inspection. The scope advances through the cannula. The internal aspect of the pedicle is visible.
Video 1B. Endoscopic pedicle inspection of an intact pedicle.
Video 2A: Endoscopic view of a medial pedicle wall breach. The video starts with the scope inside of the breach and immersed in epidural fat. The scope is then withdrawn and the defect is visible.
Video 2B: Endoscopic view of a medial pedicle wall breach with a thirty degree endoscope. The endoscope offers a more direct inspection of each wall.
Video 3: Endoscopic view of anterior pedicle breach.
Video 4A: Endoscopic view of lateral pedicle wall breach. The scope is introduced into the cannula. At 00:07, the paraspinous muscles are visible. The scope is withdrawn.
Video 4B: Endoscopic view of lateral pedicle wall breach with thirty degree scope.
Video 5: Endoscopic guided ball-tip probe inspection of the pedicle. The ball-tip probe is located at the medial defect of the pedicle. The epidural fat is visible with the endoscope. At 00:24 the ball is passed out of the pedicle into the spinal canal. The shaft of the probe is visible adjacent to the epidural fat.
Video 6: Endoscopic placement inspection of guidewire position. The guidewire is inserted in the anterior defect under careful guidance.
Video 7: Endoscopic view of a pedicle with two divergent paths. One path is at approximately 2 o’clock and one path is at 7 o’clock.
Video 8A: Endoscopic view of a pedicle with intact walls. At 00:01 the irrigation was ceased. Blood flow immediately resumed. At 00:15 irrigation was resumed. The blood was quickly flushed out and the pedicle cleared.
Video 8B: Endoscopic view of a pedicle with intact walls. The irrigation was ceased at 00:01. Blood flow immediately resumed. At 00:29 irrigation was resumed. The blood was quickly flushed out and the pedicle cleared.
Kristen Radcliff, MD, Associate Professor, Department of Orthopedic Surgery, Thomas Jefferson University, Rothman Institute, 2500 English Creek Avenue, Egg Harbor, NJ 08234.
The study was sponsored by Orthopedic Sciences, Inc. One of the authors receives compensation in the form of royalties (KR). One of the authors is an employee of the company (JB). No compensation was provided for manuscript preparation.