Development and Testing of a Novel Stabilizer Clamp to Reduce Motion During Instrumented Cervical Fusions: The Caption System

Abstract

Background

Navigation of spinal instrumentation is heavily dependent on maintaining an accurate registration of the local anatomy to the navigation software. Any motion of the anatomy reduces accuracy. Historically, navigation in the cervical spine has been limited by the mobile nature of the anatomy. Here, we describe the development and testing of a novel stabilizer clamp designed to reduce motion and improve stability for navigated-instrumentation accuracy.

Methods

Testing was completed both on a flat-top table and separately on a Jackson table. On both tables, the cadaveric specimen was positioned prone with a Mayfield clamp. A total of 30 trials were completed as 3 repeats over 10 applications with force in the anteroposterior (AP) direction, and 30 additional trials with force in the medial-lateral (ML) direction. Displacement was recorded using a digital motion indicator for each trial. The experiment was then repeated after the addition application of a novel cranial stabilization system (Caption device).

Results

The Caption cranial stabilizer significantly reduced both AP motion (45%–53% reduction) and ML motion (69%–92% reduction; P < 0.05 for each). The result was consistent when the cadaver was positioned both on the flat-top table (50.3% AP and 79.6% ML reduction) and on the Jackson table (53.5% AP and 92.5% ML reduction). With the device attached to the extended bed rails, the Caption cervical stabilization system reduced deflections by 45% AP and 69% ML when compared to deflections without the use of the stabilization system.

Conclusions

The addition of a supplementary cranial stabilization system significantly reduces cranial motion compared with the use of only a Mayfield clamp.

Clinical Relevance

Increased stability of the cranial and cervical anatomy may allow for a reduction in malposition of navigated cervical instrumentation. Particular benefit may be seen in minimally invasive cases where the bony anatomy cannot be directly visualized, and in placement of cervical pedicle screws where the anatomic tolerances required are very high.

Level of Evidence

5.

Introduction

The placement of posterior cervical screw fixation has evolved considerably over time. The most commonly used technique has involved placement of lateral mass screws with free-hand anatomic or fluoroscopic technique, which has a long history and excellent safety profile.^1–3 However, the cancellous nature of the lateral mass results in limited biomechanical strength.^3–5 In addition, the medial-to-lateral trajectory of the screw results in a medial tulip head that makes passage of a rod more difficult in constructs where both pedicle and lateral mass screws are nearby (Figure 1) and can interfere with central decompression of the canal after the implant is placed.²

Figure 1.

Caption cervical stabilization system (top, and red arrows in bottom image), and representative setup after attachment to a Mayfield 3-point head clamp (bottom).

To counter these concerns, surgeons have increasingly turned to placing cervical pedicle screws, and more recently to pedicle inlet screws, both of which have significantly improved biomechanical performance and a more lateral tulip position compared with lateral mass screws.⁵ However, the cervical pedicles are often quite narrow, and cortical violation by a pedicle screw can result in catastrophic neurological or vascular injury.^1–7 The pedicle inlet trajectory lessens this risk somewhat by stopping in the dense bone of the pedicle inlet.⁵ Nonetheless, due to the tight anatomic tolerances in this region, many authors have recommended navigating both the pedicle and pedicle inlet screws in order to minimize the risk of cortical violation.^1–7

Computed tomography-based navigation systems have been shown to improve the accuracy of spinal instrumentation.⁸ But even with this technology, screw malposition can occur, typically due to errors in anatomic registration caused by motion.⁹ Navigation accuracy requires strict registration of the regional anatomy to a local fiduciary, which is attached to the spinous processes or a skull clamp. Any motion of the cervical segments can result in discrepancies between the navigation display and the actual position of the cervical anatomy. To help minimize motion, we recently assessed a novel cervical stabilization system as a supplement to a Mayfield 3-point skull clamp (Figure 1). The purpose of the current study is to evaluate its effectiveness in reducing cranial motion as a proxy to reducing cervical motion in a cadaveric model.

Methods

The cervical stabilization system (Caption cervical stabilization system, Medtronic Inc., Minneapolis, MN) consists of 2 articulating arms that extend from bed rail attachments and connect to the sides of the Mayfield 3-point skull clamp (Integra Lifesciences Corp., Princeton, NJ; Figure 1). Testing was completed both on a standard flat-top table (Figure 2) and separately on a Jackson table (Figures 3 and 4). On both tables, a cadaveric specimen was positioned prone with a Mayfield skull clamp applied at 60N of force in a standard fashion.

Figure 2.

Test setup with the standard flat-top table. Digital force gauges were placed to detect both medial-lateral and anterior-posterior motion.

Figure 3.

Jackson table “worst-case” setup. The bed attachment comes from caudal to the patient’s shoulder and requires a rail extension (red arrows), which is an added connection and decreases the rigidity of the assembly.

To test displacement in a standardized way, a digital force gauge was used to apply 40N to the posterior skull. This magnitude of force was chosen as a surrogate for the maximum expected force a surgeon might reasonably apply when inserting posterior cervical instrumentation. A total of 30 trials were completed as 3 repeats over 10 applications with force in the anteroposterior (AP) direction and 30 additional trials with force in the medial-lateral (ML) direction. Displacement was recorded using a digital motion indicator for each trial.

The experiment was then repeated after the application of the Caption cervical stabilization system. With the Jackson table, a “worst-case” condition was simulated in which the clamps were attached to the bed caudal to the chest support (Figure 3). This position places the Caption stabilization system further from the head and requires an additional rail extension, which decreases rigidity. We also tested on a Jackson 4-poster table in what we termed a “standard” setup, which consisted of placing the Caption system cephalad to the skull. This set-up allowed the stabilizer arms to be directly attached to the bed-rail, thus eliminating the need for the additional extension piece (Figure 4). On the standard flat-top table, the bed rails are located caudal to the skull by necessity since a standard table does not extend cephalad beyond the body; therefore, only 1 Caption position was tested on the flat-top table.

Figure 4.

Test setup with Jackson 4-poster bed and the cervical stabilization system (Caption) in the standard position cephalad to the patient’s head. Because the system is closer to the head, no rail extension is needed, which increases the overall stability.

For statistical analysis, multiple comparisons (Tukey’s pairwise) were assessed after 2-way analysis of variance with stabilization and force direction as fixed factors for each table type. Significance was set with α = 0.05.

Results

Flat-Top Table

While positioned on a flat-top table without the Caption cervical stabilization system(Figure 2), application of 40N of force resulted in a mean (SD) of 1.48 (0.19) mm displacement in the AP direction and 4.3 (0.3) mm the ML direction. Motion with the Caption cervical stabilization system applied under the same testing conditions resulted in a mean (SD) of 0.74 (0.11) mm in the AP direction and 0.88 (0.2) mm in the ML direction. These findings represented a 50.3% reduction in AP and a 79.6% reduction in ML motion. Both comparisons were statistically significant (P < 0.05; Table).

Table.

Testing results.

Table Type and Stabilizer Condition	AP Deflection, mm (Mean ± SD)	Δ, Mean Reduction, %	ML Deflection, mm (Mean ± SD)	Δ, Mean Reduction (%)
Flat top without cranial stabilizer (Mayfield clamp alone)	1.48 ± 0.19	-	4.30 ± 0.30	-
Flat top with Caption cervical stabilization system	0.74 ± 0.11 a	−50.3%	0.88 ± 0.20 a	−79.6%
Jackson without cervical stabilization system (Mayfield clamp alone)	0.55 ± 0.03	-	6.23 ± 1.02	-
Jackson with worst-case Caption cervical stabilization system	0.30 ± 0.02 a	−45.0%	1.92 ± 0.23 a	−69.2%
Jackson with standard position Caption cervical stabilization system	0.26 ± 0.02 a	−53.5%	0.50 ± 0.14 a	−92.0%

Abbreviation: ML, medial-lateral.

Note: Caption cervical stabilization system from Medtronic Inc. (Minneapolis, MN, USA). Δ = reduction from the no cranial stabilizer (Mayfield clamp alone) set-up.

^aSignificant reduction (P < 0.01).

Jackson Table Standard Positioning

While positioned on a 4-post Jackson table without the Caption cervical stabilization system, application of 40N of force resulted in a mean (SD) of 0.55 (0.03) mm in the AP direction and 6.23 (1.02) mm the ML direction. With the cervical stabilization system applied in the standard position (Figure 4), mean (SD) motion was reduced to 0.26 (0.02) mm the AP direction and 0.5 (0.14) mm the ML direction. These findings represented a 53.5% reduction in AP motion and a 92% reduction in ML motion. Both comparisons were statistically significant (P < 0.05; Table).

Jackson Table Worst-Case Positioning

To simulate a worst-case setup, the Caption cervical stabilization system was attached to the extended bed rails caudal to the patient’s shoulders (Figure 3). In this position, application of 40N of force resulted in a mean (SD) of 0.3 (0.02) mm motion in the AP direction and a mean (SD) of 1.92 (0.23) mm in the ML direction. These findings represented a 45% reduction in AP motion and a 69.2% reduction in ML motion, as compared with the measurements made without the stabilizer. Both comparisons were statistically significant (P < 0.05; Table).

Discussion

Spinal motion that disrupts the registration of anatomy to the reference frame is a significant source of inaccuracy when placing navigated posterior spinal instrumentation.¹⁰ This concern is particularly acute in the cervical spine, where malpositioned hardware can have catastrophic consequences due to the close proximity of the vertebral artery and exiting nerve roots of the brachial plexus.⁸ To minimize navigation errors associated with spinal motion, we developed a supplemental cercvical stabilization system that attaches to the Mayfield 3-point skull clamp. As described in the present study, the application of the device resulted in a significant reduction in cranial motion in a cadaveric study. Several of our findings merit further discussion.

The efficacy of the cervical stabilization system in reducing cranial motion was tested in 3 different bed positions (Table). In each case, cranial motion was significantly reduced with application of the cervical stabilization system, both in the AP (45%–53% reduction) and ML directions (69%–92% reduction; P < 0.05 for each). The magnitude of reduction in motion ranged from 0.2 to 6.2 mm (Table). Given the tight anatomic tolerances in the cervical spine, where cervical pedicles are often less than 3 to 4 mm in size, we believe this reduction in motion is clinically relevant. The cervical stabilization system was most effective when positioned without the rail extender and attached directly to the bed clamps (Figure 4). When positioned caudal to the patient’s head, the system requires a longer bed-rail attachment (Figure 3), which reduces the rigidity of the construct and results in less overall stabilization. Some patients’ anatomy and some bed types may require the use of the rail extenders. However, where possible, we recommend surgeons apply the clamp without the rail extenders, as shown in Figure 4.

Notably, all imaging was done with an O-Arm passed around the patient, and the 2 clamps of the cervical stabilization system did not interfere with O-Arm passage around the patient in any position. While the skull clamps do increase the total width of the Mayfield complex, they pass below the axis of the spine and thus do not create radiographic interference on imaging. Therefore, we believe the cervical stabilization system should not interfere with standard workflows for implant placement.

The cervical stabilization system attach to both sides of the Mayfield 3-point skull clamp, and, given this design, it was more effective at reducing ML motion than AP motion. Reduction in ML motion is most critical for navigation accuracy in the cervical spine because medial breaches can result in impingement into the canal or neuroforamen. Additional improvements in AP rigidity could be achieved with the application of a third clamp that is attached to the bottom of the Mayfield 3-point skull clamp. However, the increased bulk of the Mayfield clamp makes it more difficult for imaging to pass around the patient and increases the complexity of the Mayfield clamp application. Furthermore, even with the Mayfield skull clamp alone, the magnitude of AP motion was relatively small (0.5–1.48 mm), which is in contrast to the ML direction (4.3–6.2 mm), which was more mobile with the skull clamp alone. Application of the cervical stabilization system reduced the AP motion to very small amounts (0.26–0.74 mm). So, even though the percentage reduction in AP motion appears less than that achieved in the ML direction, the absolute magnitude of residual AP motion remains low, and this reflects in part the fact that there was less AP motion to begin with.

Our study does have several weaknesses. Our study used cranial motion as a proxy for overall cervical motion. Previous studies have shown that cranial motion and cervical motion are closely correlated,^11,12 and thus, we believe that a significant reduction in cranial motion should also translate into a significant reduction in cervical motion. Nonetheless, cervical motion is not eliminated by this device. Surgeons should take care to minimize the amount of force used when placing instrumentation in order to limit navigation errors caused by spinal motion. Second, given our study design, we cannot conclude definitively whether or not the reduced cranial motion translates into improved navigation accuracy. A separate cadaveric study is planned to assess this possibility.

Conclusion

Overall, the addition of the supplementary Caption cervical stabilization system significantly reduced cranial motion compared with the use of a Mayfield 3-point skull clamp alone. These preliminary cadaveric studies are encouraging, and both future cadaveric and clinical studies are planned to determine whether the increased stability leads to a reduction in malposition of navigated cervical instrumentation. Particular benefit may be seen in minimally invasive cases where the bony anatomy cannot be directly visualized and in the placement of cervical pedicle screws where the anatomic tolerances required are tight.