Lesions of the ventral epidural space of the upper cervical spinal cord pose a unique challenge. The transoral transpharyngeal approach has historically been described as an effective approach to these pathologies,1–7 but risks such as infection, cerebrospinal fluid fistula, and postoperative pulmonary complications have resulted in high morbidity and mortality, leading surgeons to consider alternative approaches to this region.8–10 Although the endoscopic endonasal approach has advantages over the transoral approach, resection of the odontoid process and transverse ligament requires posterior fixation and fusion. Endoscopic endonasal approaches also require a skilled multidisciplinary team, which may not be readily available at all spine centers. The direct lateral approach (also known as the far lateral transatlas, extreme lateral transatlas, or anterolateral approach) involves C1 hemilaminectomy and partial or complete lateral mass resection and has been proposed as one alternative.11,12
Although some ventral atlantoaxial pathologies are the result of chronic instability and require fusion for definitive treatment, some—such as tumors, epidural abscesses, and epidural hematomas—only require access to the ventral pathology without the need for stabilization. In these scenarios, the direct lateral transatlas approach may best accomplish this goal. When complete lateral mass resection is performed, this leads to clinical instability and instrumented stabilization is required.11,13–15 To date, however, no biomechanical studies have been performed to evaluate occipitocervical stability after partial C1 lateral mass resection. Studies of condylectomy with the far lateral approach suggest that 50% of the condyle can be resected without significant increases in range of motion (ROM),16 but these studies are not directly applicable to the direct lateral transatlas approach.
In this study, we performed an in vitro biomechanical evaluation of the craniovertebral junction after sequential unilateral C1 lateral mass resection. We hypothesized that partial resection of the lateral mass would not result in a significant increase in ROM, and may therefore not require internal stabilization.
Methods
Cadaveric Specimens
We obtained eight fresh-frozen, unembalmed human cadaveric cephali consisting of the occipital bone (Oc) to at least C3. Donors were between 18 and 75 years of age with a body mass index between 18.5 and 40 at the time of death. Donors with a history of craniovertebral surgery, spinal malignancy, or rheumatoid disease were excluded. Occipitocervical radiographs were also reviewed to rule out traumatic injuries, spinal fusions, and craniocervical deformities. The craniums were transected rostrally from just above the supraorbital rim to the inion, and all soft tissue and musculature was removed while preserving ligaments and joint capsules. Specimens were stored at −20°C and thawed to room temperature for testing. Specimens were thawed once, and all testing for each specimen was performed in a single session.
Biomechanical Testing
We performed in vitro ROM tests using a 6-axis robotic spine testing system (KR16, Kuka Robotics) with a 6-axis force-moment sensor (Delta/SI-330-30, ATI Industrial Automation) to measure applied loads (Fig. 1). An optoelectronic camera system (Optotrak, Northern Digital, Inc.) captured segmental ROM data. The robotic system was controlled using simVITRO software (Cleveland Clinic) in order to apply pure moments to the spine.
Robot and cadaver testing setup with optoelectronic sensors (black arrow) and load cell (white arrow). Figure is available in color online only.
Specimens were rigidly affixed to the robotic testing system at the Oc and C3, and infrared optoelectronic sensors were rigidly fixed to C1 and C2 without causing any bony destruction, as previously published.16 The spatial relationships between the robot, load cell, and each vertebral segment (Oc–C3) were established using a digitizer made up of motion tracking markers. Coordinate system definitions for the vertebral segments were established according to Techy et al.17 Prior to testing, each specimen was preconditioned with three loading and unloading cycles of ±1.5 Nm along flexion/extension (FE), axial rotation (AR), and lateral bending (LB) to minimize viscoelastic effects. The robot applied a constant compressive force of 40 N throughout testing to simulate head weight load.
In each testing state, the robot applied ±1.5 Nm of torque of in FE, AR, and LB. The robot also simulated coupled ROM by applying ±1.5 Nm of torque in AR after applying 1.5 Nm of torque in extension (AR+E) and flexion (AR+F). Similarly, coupled ROM was performed by applying ±1.5 of Nm torque in LB after applying 1.5 Nm of torque in the left AR (LAR; LB+LAR) and right AR (RAR; LB+RAR). In each testing state, ROM testing was performed in the same order, with the coupled ROM testing first, followed by LB, FE, and AR, for a total of 14 loading trajectories. During each loading trajectory, primary and off-axis loads, along with rotational rate of change, were monitored. The loading condition was terminated when each load and rotational rate were within a prescribed range of their target value (±2, ±5, and ±2 N for posterior, compressive, and lateral shear load, respectively; ±0.1 Nm for F, AR, and LB torque; and ±0.1°/sec for F, AR, and LB rate of change). Overall spine, Oc–C1, and C1–2 ROM were recorded for each loading trajectory. The specimens were returned to a predefined visual and force-neutral position between each loading trajectory.
Testing Conditions
The specimens were first tested in the native state to collect baseline ROM values. ROM testing was then performed after each of the following sequential surgical conditions: C1 hemilaminectomy, 25% lateral mass resection, 50% lateral mass resection, 75% lateral mass resection, and 100% lateral mass resection. The surgical procedure was performed on the right side for all specimens. Lateral mass resection was performed in a posterior-to-anterior direction (Fig. 2), and care was taken to preserve all ligamentous attachments except as needed for bony resection. In order to perform the C1 hemilaminectomy, the posterior atlantooccipital and atlantoaxial ligaments were detached unilaterally from the posterior arch of C1 from midline to the posterior margin of the C1 lateral mass. After the C1 hemilaminectomy was complete, the posterior and anterior walls of the lateral mass were digitized to define the 0% and 100% margins for lateral mass resection. A LabVIEW (National Instruments) custom program was used to capture the digitizer input in real time and identify the location of each sequential lateral mass resection relative to these margins. The Oc–C1 and C1–2 joint capsules were disrupted in a posterior-to-anterior direction, only as needed to perform resection of the C1 lateral mass. The anterior atlantooccipital and atlantoaxial ligaments were violated only at 100% lateral mass resection as was necessary for complete lateral mass resection. The cruciate, alar, and apical ligaments were not violated in the surgical procedure.
Right lateral view of C1 showing the lateral mass divided into quartiles for sequential resection. OC = occipital condyle; LM = C1 lateral mass; PA = C1 posterior arch. Figure is available in color online only.
Statistical Analysis
Overall ROMs in each primary movement (FE, AR, and LB) and in each coupled movement (AR+E, AR+F, LB+LAR, and LB+RAR) were compared among each testing condition using two-way repeated-measures ANOVA to allow each specimen to serve as its own control to account for baseline differences in ROM. Post hoc Tukey-Kramer analysis was performed for multiple comparisons between groups. All analyses were performed using RStudio, and p < 0.05 was defined as statistically significant.
Results
Testing Trajectories
The prescribed loading conditions were accurately applied. Across all 672 testing trajectories (for all specimens and surgical conditions), the root mean square errors between desired and actual loads were, on average, 0.3 ± 0.9 N for posterior shear, 1.4 ± 1.2 N for compression, 0.2 ± 0.3 N for lateral shear, 0.03 ± 0.03 Nm for F torque, 0.02 ± 0.02 Nm for AR torque, and 0.05 ± 0.05 Nm for LB torque.
Oc–C1 Range of Motion
In the intact state, mean ROMs for AR, FE, and LB were 7.9° ± 3.0°, 26.0° ± 4.7°, and 6.9° ± 3.1°, respectively. Mean coupled Oc–C1 ROMs for AR+E, AR+F, LB+LAR, and LB+RAR were 7.3° ± 3.9°, 3.9° ± 1.2°, 4.8° ± 2.1°, and 4.4° ± 2.1°, respectively. Right-sided C1 hemilaminectomy did not significantly increase Oc–C1 ROM in any primary or coupled movement (Table 1).
Oc–C1 primary and coupled ROM in intact and surgical states
| Movement | Intact | C1 | Lateral Mass Resection | |||
|---|---|---|---|---|---|---|
| 25% | 50% | 75% | 100% | |||
| AR | 7.9° ± 3.0° | 9.1° ± 3.0° 15.2% (0.866) | 10.5° ± 3.9° 32.9% (0.156) | 12.2° ± 2.9° 54.4% (0.002) | 14.0° ± 4.2° 77.2% (<0.001) | 17.0° ± 6.5° 115.2% (<0.001) |
| FE | 26.0° ± 4.6° | 26.6° ± 4.7° 2.3% (0.986) | 26.4° ± 4.3° 1.5% (0.998) | 27.4° ± 4.9° 5.4% (0.749) | 27.1° ± 5.0° 4.2% (0.883) | 26.6° ± 5.5° 2.3% (0.991) |
| LB | 6.9° ± 3.1° | 7.5° ± 3.1° 8.7% (0.984) | 7.8° ± 3.2° 13.0% (0.927) | 10.2° ± 2.2° 47.8% (0.010) | 12.7° ± 2.5° 84.1% (<0.001) | 13.8° ± 4.2° 100.0% (<0.001) |
| AR+E | 7.3° ± 3.9° | 8.7° ± 4.2° 19.2% (0.746) | 9.9° ± 4.6° 35.6% (0.169) | 12.1° ± 3.3° 65.8% (<0.001) | 13.1° ± 4.4° 79.5% (<0.001) | 15.6° ± 7.2° 113.7% (<0.001) |
| AR+F | 3.9° ± 1.2° | 4.6° ± 1.7° 17.9% (0.972) | 4.9° ± 1.7° 25.6% (0.875) | 5.6° ± 1.4° 43.6% (0.446) | 6.3° ± 1.5° 61.5% (0.104) | 9.3° ± 4.3° 138.5% (<0.001) |
| LB+LAR | 4.8° ± 2.1° | 4.8° ± 1.9° 0.0% (>0.999) | 5.2° ± 2.0° 8.3% (0.999) | 6.1° ± 1.9° 27.0% (0.744) | 9.4° ± 5.2° 95.8% (<0.001) | 10.5° ± 4.6° 118.8% (<0.001) |
| LB+RAR | 4.4° ± 2.1° | 4.6° ± 2.2° 4.5% (>0.999) | 5.0° ± 2.8° 13.6% (0.943) | 6.0° ± 2.9° 36.4% (0.216) | 7.6° ± 3.9° 72.7% (<0.001) | 9.0° ± 5.1° 104.5% (<0.001) |
Values presented as mean ± SD (first line) and percentage change in ROM from intact and associated p values in parentheses (second line). Boldface type indicates statistical significance.
Compared with the intact state, 25% lateral mass resection resulted in no significant change in ROM for any primary or coupled movement. Fifty percent lateral mass resection, compared with the intact state, resulted in a significant increase in ROM for AR (4.3°, p = 0.002), LB (3.3°, p = 0.010), and AR+E (4.8°, p < 0.001). Compared with the intact state, 75% and 100% resection resulted in further significant increases in the Oc–C1 ROM in all movements except FE.
C1–2 Range of Motion
In the intact state, mean ROMs for AR, FE, and LB were 63.2° ± 20.1°, 17.0° ± 5.3°, and 7.3° ± 6.5°, respectively. Mean coupled C1–2 ROMs for AR+E, AR+F, LB+LAR, and LB+RAR were 58.6° ± 19.8°, 60.6° ± 19.6°, 4.2° ± 2.8°, and 4.1° ± 2.8°, respectively. Right-sided C1 hemilaminectomy did not significantly increase C1–2 ROM in any primary or coupled movement (Table 2).
C1–2 primary and coupled ROM in intact and surgical states
| Movement | Intact | C1 | Lateral Mass Resection | |||
|---|---|---|---|---|---|---|
| 25% | 50% | 75% | 100% | |||
| AR | 63.2° ± 20.1° | 64.6° ± 19.5° 2.2% (0.999) | 65.2° ± 18.4° 3.2% (0.993) | 66.5° ± 17.7° 5.2% (0.928) | 68.5° ± 16.9° 8.4% (0.652) | 73.0° ± 19.0° 15.5% (0.075) |
| FE | 17.0° ± 5.3° | 18.0° ± 5.6° 5.9% (0.968) | 18.5° ± 5.6° 8.8% (0.832) | 18.7° ± 6.3° 10.0% (0.725) | 21.5° ± 7.5° 26.5% (0.008) | 26.4° ± 9.7° 55.3% (<0.001) |
| LB | 7.3° ± 6.5° | 7.4° ± 6.6° 1.4% (>0.999) | 7.8° ± 6.7° 6.8% (>0.999) | 9.1° ± 7.1° 24.7% (0.980) | 13.5° ± 9.2° 84.9% (0.178) | 22.7° ± 13.6° 211.0% (<0.001) |
| AR+E | 58.6° ± 19.8° | 59.2° ± 19.3° 1.0% (>0.999) | 59.3° ± 17.9° 1.2% (>0.999) | 59.4° ± 15.8° 1.4% (>0.999) | 61.7° ± 14.5° 5.3% (0.986) | 66.7° ± 20.3° 13.8% (0.540) |
| AR+F | 60.6° ± 19.6° | 61.6° ± 19.1° 1.7% (0.981) | 62.5° ± 18.4° 3.1% (0.799) | 63.4° ± 17.8° 4.6% (0.434) | 64.5° ± 16.6° 6.4% (0.110) | 63.4° ± 17.1° 4.6% (0.407) |
| LB+LAR | 4.2° ± 2.8° | 4.1° ± 2.7° −2.4% (>0.999) | 4.2° ± 2.8° 0.0% (>0.999) | 4.8° ± 3.2° 14.3% (0.912) | 6.1° ± 3.8° 45.2% (0.013) | 7.4° ± 3.9° 76.2% (<0.001) |
| LB+RAR | 4.1° ± 2.8° | 4.2° ± 2.7° 2.4% (>0.999) | 4.4° ± 2.6° 7.3% (>0.999) | 5.2° ± 3.0° 26.8% (0.846) | 6.0° ± 3.5° 46.3% (0.306) | 10.0° ± 6.0° 143.9% (<0.001) |
Values presented as mean ± SD (first line) and percentage change in ROM from intact and associated p values in parentheses (second line). Bold typeface indicates statistical significance.
Compared with the intact state, 25% and 50% lateral mass resection resulted in no significant change in ROM for any primary or coupled movement. Seventy-five percent lateral mass resection, compared with the intact state, resulted in a significant increase in ROM for FE (4.5°, p = 0.008) and LB+LAR (1.9°, p = 0.013), respectively. Complete resection resulted in further increases in the C1–2 ROM in all movements except AR+F.
Oc–C2 Range of Motion
In the intact state, mean ROMs for AR, FE, and LB were 71.2° ± 18.4°, 43.0° ± 9.2°, and 14.2° ± 9.2°, respectively. Mean coupled Oc–C2 ROMs for AR+E, AR+F, LB+LAR, and LB+RAR were 65.8° ± 17.6°, 64.5° ± 19.7°, 9.0° ± 4.4°, and 8.5° ± 4.0°, respectively. Right-sided C1 hemilaminectomy did not significantly increase Oc–C2 ROM in any primary or coupled movement (Table 3).
Oc–C2 primary and coupled ROM in intact and surgical states
| Movement | Intact | C1 | Lateral Mass Resection | |||
|---|---|---|---|---|---|---|
| 25% | 50% | 75% | 100% | |||
| AR | 71.2° ± 18.4° | 73.7° ± 18.4° 3.5% (0.988) | 75.6° ± 17.9° 6.2% (0.874) | 78.8° ± 17.4° 10.7% (0.426) | 82.5° ± 17.8° 15.6% (0.077) | 90.0° ± 22.9° 26.4% (<0.001) |
| FE | 43.0° ± 9.2° | 44.6° ± 9.6° 3.7% (0.468) | 44.8° ± 9.2° 4.2% (0.319) | 46.1° ± 9.6° 7.2% (0.016) | 48.6° ± 9.9° 13.0% (<0.001) | 53.0° ± 11.1° 23.3% (<0.001) |
| LB | 14.2° ± 9.2° | 15.0° ± 9.3° 5.6% (>0.999) | 15.6° ± 9.4° 9.9% (0.996) | 19.3° ± 8.8° 35.9% (0.452) | 26.2° ± 11.2° 84.5% (0.001) | 36.5° ± 15.6° 157.0% (<0.001) |
| AR+E | 65.8° ± 17.6° | 67.9° ± 17.7° 3.2% (0.999) | 69.1° ± 16.8° 5.0% (0.988) | 71.5° ± 15.9° 8.7% (0.885) | 74.8° ± 16.5° 13.7% (0.541) | 82.3° ± 26.7° 25.1% (0.037) |
| AR+F | 64.5° ± 19.7° | 66.2° ± 19.6° 2.6% (0.929) | 67.4° ± 18.9° 4.5% (0.604) | 68.9° ± 18.1° 6.8% (0.161) | 70.8° ± 17.2° 9.8% (0.014) | 72.7° ± 18.6° 12.7% (0.001) |
| LB+LAR | 9.0° ± 4.4° | 9.0° ± 4.3° 0.0% (>0.999) | 9.3° ± 4.5° 3.3% (>0.999) | 10.9° ± 4.8° 21.1% (0.763) | 15.6° ± 8.5° 73.3% (<0.001) | 17.9 °± 7.9° 98.9% (<0.001) |
| LB+RAR | 8.5° ± 4.0° | 8.8° ± 3.9° 3.5% (>0.999) | 9.4° ± 4.4° 10.6% (0.985) | 11.1° ± 5.0° 30.6% (0.392) | 13.6° ± 6.0° 60.0% (0.008) | 19.0° ± 9.2° 123.5% (<0.001) |
Values presented as mean ± SD (first line) and percentage change in ROM from intact and associated p values in parentheses (second line). Bold typeface indicates statistical significance.
Compared with the intact state, 25% lateral mass resection resulted in no significant change in ROM for any primary or coupled movement. Fifty percent lateral mass resection, compared with the intact state, resulted in a significant increase in FE ROM of 3.1° (p = 0.016). Compared with the intact state, 75% and 100% resection resulted in further increases in the Oc–C2 ROM in all movements (Fig. 3).
Bar graph showing Oc–C2 ROM in each testing state (error bars represent standard error of the mean). Asterisks denote statistically significant increases in ROM, compared with the intact state.
Discussion
In this study, we investigated the biomechanical stability of the craniovertebral junction after C1 hemilaminectomy and sequential lateral mass resection and found that 50% resection of the lateral mass resulted in significant increases in Oc–C1, C1–2, and Oc–C2 ROM. This is the first study to evaluate the biomechanics of lateral mass resection, and our results provide meaningful insight for the direct lateral transatlas approach. This approach offers distinct advantages over more traditional approaches. It acts as a caudal extension of the far lateral approach, with better ventral access to the C1–2 region. Additionally, unlike the far lateral approach, the direct lateral approach is purely extradural and does not require extensive opening of the posterior fossa. Given these advantages and our findings, the direct lateral approach offers unique and clinically relevant applications. We have previously published our step-by-step surgical technique for the direct lateral approach.11
In 2002, Türe and Pamir discussed a direct lateral transatlas approach with resection of the lateral mass for access to the ventral epidural space with concurrent unilateral posterior occipitocervical fixation via the same incision.12 This obviated the need for a second operation for stabilization and avoided many of the risks associated with anterior transoral or transnasal approaches. Abdullah et al.11 described uses of the direct lateral approach, choosing to perform occipitocervical fusion based on the surgeon’s clinical judgment of postoperative instability, whereas Adada et al.18 performed prophylactic occipitocervical fusion in all direct lateral approaches. Our results suggest that for certain pathologies, occipitocervical fusion may not be required when 25% or less of the lateral mass is resected.
In a prior biomechanical study by Kshettry et al. of occipitocervical stability with sequential occipital condyle resection, significant increases in ROM were only achieved after 75% resection of the condyle.16 In contrast, we found that 50% resection of the C1 lateral mass led to significant increases in ROM. This is likely because the C1 lateral mass functions as a two-joint structure (Oc–C1 and C1–2), whereas the occipital condyle is only a one-joint structure. Additionally, Kshettry et al. performed joint-sparing condylectomy, whereas we disrupted both the Oc–C1 and C1–2 joint spaces in this study. For certain ventral C2 pathologies, if adequate surgical access can be achieved by removing lateral mass while preserving the Oc–C1 joint, it may be possible to retain greater native stability.
In our clinical experience, we have found that less than 25% lateral mass resection is generally required for ventral epidural pathologies such as hematoma or abscess. Although some ventral epidural tumors may require additional resection, this can typically still be less than 50%. In Fig. 4, we show a postoperative axial CT after a right-sided direct lateral approach in which less than 25% of the lateral mass was resected. In this case, we were afforded adequate working space, and the patient did not show any signs of instability.
Postoperative axial CT after right-sided direct lateral approach showing less than 25% C1 lateral mass resection. The black lines show the expanded trajectory toward the ventral epidural space.
The results of this study should be interpreted within the context of its limitations. Although we used a validated biomechanical method in this study, in nonhealing cadaveric models only testing for acute instability can be performed, which does not address the effects of repeated cyclical loading and unloading (i.e., chronic instability). However, the ROMs for the intact specimens in our study were similar to previously reported values, which adds to the reproducibility and precision of our methodology.13,16,19 We also chose to evaluate the biomechanical effects of non–joint-sparing lateral mass resection, but in certain clinical scenarios, the joint may be preserved, which could lead to a lesser degree of instability with lateral mass resection. In this study, we chose predefined surgical conditions, which limits our ability to draw conclusions for lateral mass resections between 25% and 50%. Finally, we used increase in ROM as a surrogate for instability, but there are no precise ROM definitions for occipitocervical instability, and clinical factors should be included in decision-making in this patient population.
Conclusions
In this cadaveric biomechanical study, we found that 25% resection of the C1 lateral mass did not result in significant biomechanical instability at the occipitocervical junction, and 50% resection led to significant increases in Oc–C2 ROM. This is the first biomechanical study of lateral mass resection, and future studies can serve to validate these findings.
Acknowledgments
This study was funded by the Cleveland Clinic Research Program Committee (RPC #1389).
Disclosures
Callan M. Gillespie: received royalties from CCF/simVITRO. Robb W. Colbrunn: received royalties from simVITRO-Cleveland Clinic. Michael P. Steinmetz: received royalties from Zimmer/Biomet and Elsevier; consultant for Globus and Stryker; received honoraria from Medtronic and Globus. Varun R. Kshettry: consultant for Integra and Stryker.
Author Contributions
Conception and design: Kshettry, Soni, Loss, Gillespie, Colbrunn, Benzel. Acquisition of data: Soni, Loss, Gillespie. Analysis and interpretation of data: Kshettry, Soni, Loss, Gillespie, Colbrunn, Recinos. Drafting the article: Soni. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Statistical analysis: Soni. Administrative/technical/material support: Loss, Gillespie, Steinmetz. Study supervision: Kshettry, Colbrunn, Schlenk, Steinmetz, Recinos, Benzel.
References
- 1
Joshi K, Woodard T, Borghei-Razavi H, et al. Endoscopic endonasal odontoidectomy. J Neurol Surg B Skull Base. 2019;80(4)(suppl 4):S370.
- 2
Mazzatenta D, Zoli M, Mascari C, et al. Endoscopic endonasal odontoidectomy: clinical series. Spine (Phila Pa 1976).2014;39(10):846–853.
- 3
Morales-Valero SF, Serchi E, Zoli M, et al. Endoscopic endonasal approach for craniovertebral junction pathology: a review of the literature. Neurosurg Focus. 2015;38(4):E15.
- 4
Ogiwara T, Miyaoka Y, Nakamura T, et al. Endoscopic endonasal odontoidectomy in the hybrid operating room. World Neurosurg. 2019;131:137–140.
- 5
Ruetten S, Hahn P, Oezdemir S, et al. Full-endoscopic uniportal retropharyngeal odontoidectomy for anterior craniocervical infection. Minim Invasive Ther Allied Technol. 2019;28(3):178–185.
- 6
Yu Y, Hu F, Zhang X, Sun C. Endoscopic transnasal odontoidectomy. Sports Med Arthrosc Rev. 2016;24(1):2–6.
- 7
Zoli M, Mazzatenta D, Valluzzi A, et al. Endoscopic endonasal odontoidectomy. Neurosurg Clin N Am. 2015;26(3):427–436.
- 8
Dickman CA, Crawford NR, Brantley AG, Sonntag VK. Biomechanical effects of transoral odontoidectomy. Neurosurgery. 1995;36(6):1146–1153.
- 9
Shriver MF, Kshettry VR, Sindwani R, et al. Transoral and transnasal odontoidectomy complications: a systematic review and meta-analysis. Clin Neurol Neurosurg. 2016;148:121–129.
- 10
Tubbs RS, Demerdash A, Rizk E, et al. Complications of transoral and transnasal odontoidectomy: a comprehensive review. Childs Nerv Syst. 2016;32(1):55–59.
- 11↑
Abdullah KG, Schlenk RS, Krishnaney A, et al. Direct lateral approach to pathology at the craniocervical junction: a technical note. Neurosurgery. 2012;70(2 Suppl Operative):202–208.
- 12↑
Türe U, Pamir MN. Extreme lateral-transatlas approach for resection of the dens of the axis. J Neurosurg. 2002;96(1)(suppl):73–82.
- 13↑
Panjabi MM, Oda T, Crisco JJ III, et al. Experimental study of atlas injuries. I. Biomechanical analysis of their mechanisms and fracture patterns. Spine (Phila Pa 1976).1991;16(10)(suppl):S460–S465.
- 14
Smith TJ. In vitro spinal biomechanics. Experimental methods and apparatus. Spine (Phila Pa 1976).1991;16(10):1204–1210.
- 15
Steinmetz MP, Mroz TE, Benzel EC. Craniovertebral junction: biomechanical considerations. Neurosurgery. 2010;66(3)(suppl):7–12.
- 16↑
Kshettry VR, Healy AT, Colbrunn R, et al. Biomechanical evaluation of the craniovertebral junction after unilateral joint-sparing condylectomy: implications for the far lateral approach revisited. J Neurosurg. 2017;127(4):829–836.
- 17↑
Techy F, Mageswaran P, Colbrunn RW, et al. Properties of an interspinous fixation device (ISD) in lumbar fusion constructs: a biomechanical study. Spine J. 2013;13(5):572–579.
- 18↑
Adada B, Vera Silva MA, Darwish H, Dakwar E. Far-lateral trans-atlas extradural resection of retro-odontoid synovial cyst: surgical technique and review of literature. Interdiscip Neurosurg. 2019;17:28–35.
- 19↑
Panjabi M, Dvorak J, Crisco J III, et al. Flexion, extension, and lateral bending of the upper cervical spine in response to alar ligament transections. J Spinal Disord. 1991;4(2):157–167.
