ANATOMIC REPORT

CADAVERIC MORPHOMETRIC ANALYSIS LATERAL MASS SCREW PLACEMENT

Michael Y. Wang, M.D.
Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, California

FOR

ATLANTAL

Srinath Samudrala, M.D.

Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, California Reprint requests: Michael Y. Wang, M.D., 1200 North State Street, Suite #5046, Los Angeles, CA 90033. Email: myw@usc.edu Received, September 23, 2003. Accepted, February 13, 2004.

OBJECTIVE: Atlantal lateral mass screws provide an alternative to C1/C2 transarticular screws and, in some cases, can obviate the need for extending a fusion to the occiput. For these reasons, C1 lateral mass screws are becoming increasingly popular. However, the critical local anatomy and unfamiliarity with this new technique can make C1 screw placement more challenging. METHODS: Morphometric analysis was performed on 74 cadaveric spines obtained from the Department of Anatomy at the Keck School of Medicine, University of Southern California. Critical measurements were determined for screw entry points, trajectories, and lengths for application of the technique described by Harms and Melcher. RESULTS: The mean height and width for screw entry on the posterior surface of the lateral mass were 3.9 and 7.3 mm, respectively. The maximum medialized screw trajectory ranged from 25 to 45 degrees (mean, 33 degrees). The mean maximal screw length to obtain bicortical purchase was 22.5 mm, and the mean minimum screw depth was 14.4 mm. Screw depths varied on the basis of the entry point, trajectory, and vertebral morphology. The overhang of the posterior arch averaged 11.4 mm (range, 6.917 mm). All specimens could accommodate 3.5-mm lateral mass screws bilaterally with proper preparation of the entry site. CONCLUSION: Significant variations in the morphology of C1 exist. However, the large size of the atlantal lateral mass makes screw placement forgiving. Preoperative computed tomographic scans and intraoperative fluoroscopy are useful in guiding proper screw placement. Close attention should be paid to preparation of the screw entry site.
KEY WORDS: Anatomy, Atlas, Cervical spine, Occipitocervical fusion
Neurosurgery 54:1436-1440, 2004
DOI: 10.1227/01.NEU.0000124753.74864.07

www.neurosurgery-online.com

he clinical application of atlantal lateral mass screws was first reported by Goel and Laheri in 1994 (2) but was later popularized by Harms and Melcher (7), using a polyaxial screw-rod system, in 2001. Before this report, biomechanical studies had demonstrated the superiority of Magerl C1/C2 transarticular screws over wiring for stabilization of the atlantoaxial region (5, 8, 11, 12), making Magerl screws the technique of choice when fixating this region of the spine. However, because the small size of the C2 isthmus precludes the placement of transarticular screws in up to 20% of patients (13, 15), the technique of atlantal lateral mass screw fixation was developed to overcome this limitation. During the past 2 years, C1 screws have gained tremendous popularity, and multiple case series have attested to the safety and feasibility of this technique (3, 14, 15, 17). Furthermore, biomechanical testing in both intact and destabilized cadaveric

models demonstrated that C1 lateral mass screws in conjunction with C2 pedicle screws had a similar biomechanical profile compared with Magerl transarticular screw fixation (4, 10, 16). Advantages of this new technique include the fact that the large size of the C1 lateral mass makes the procedure more forgiving than transarticular screw placement, rendering screw placement feasible in almost all patients. The steep, cranially directed trajectories used with Magerl screws are also eliminated, obviating the need for long incisions and percutaneous insertion. In addition, because screws can be placed independently at C1 and C2, sagittal and rotational atlantoaxial deformities can be reduced intraoperatively, eliminating the surgeons dependence on preoperative techniques for spinal realignment. However, the critical proximate anatomy can make C1 lateral mass screw placement challenging. The screw entry site is typically covered by a large paravertebral venous plexus,

1436 | VOLUME 54 | NUMBER 6 | JUNE 2004

www.neurosurgery-online.com

ATLANTAL LATERAL MASS SCREW ANATOMY

which can cause troublesome bleeding and obscure good visualization. Furthermore, the spinal cord, vertebral artery, and cervical nerve roots can all potentially be injured by implant misplacement. This morphometric study was undertaken to determine the bony anatomy critical for proper C1 lateral mass screw placement.

Figure 2. Screw entry points were on the posterior surface of the lateral mass below the dorsal arch of C1, and lengths were determined on the basis of a technique that used bicortical purchase. Linear measurements were made with the use of electronic calipers (Maryland Metrics, Baltimore, MD) to the nearest 0.1 mm and from cortical surface to cortical surface. Maximum screw trajectory angles were determined for 4.0mm-diameter screws to ascertain the most severe degrees of angulation that avoided screw exposure through the bone. All measurements were obtained bilaterally. Statistical analysis was performed with Analyze-It software (Leeds, England).

RESULTS
Screw entry points were identified on the posterior surface of the lateral mass beneath the insertion of the C1 posterior arch. The mean height and width were 3.9 and 7.3 mm, respectively. Forty-eight (65%) of the specimens had an entry height of less than 4 mm at the midportion of the lateral mass, and 30 of these had a height of less than 3.5 mm (41%). All lateral masses had a width greater than 3.5 mm, and only two had a width of less than 4 mm. Significant flexibility was found in terms of the medial and lateral screw trajectories that would allow good bone purchase. Maximum screw medialization ranged from 25 to 45 degrees (mean, 33 degrees). Standard screw placement does not involve a lateralized trajectory, but the larger specimens allowed up to 45 degrees of lateral deviation before the proximal screw shaft encroached on the spinal canal. Screw depths varied significantly depending on the screw entry point, trajectory, and vertebral morphology. Maximum screw depth was identified by use of a superior and medial-

MATERIALS AND METHODS

Seventy-four adult dried human cadaver spines were used in this study. Specimens were obtained from the Department of Cellular FIGURE 1. Topographic anatomy of and Neurobiology, Keck the atlas. Sup., superior; Inf., inferiSchool of Medicine, Univer- or; Ant., anterior; Post., posterior. sity of Southern California, after approval had been obtained from the Willed Body Program. All specimens were inspected to ensure that the vertebrae were intact and free of osteophytes or metastatic tumors before measurements were made. Representative dimensions were defined as in Figure 1 and Table 1 and were determined for screw placement by the technique described by Harms and Melcher (7), as shown in

TABLE 1. Measured dimensions for Figure 1 Mean A Entry point height (mm) B Entry point width (mm) C Maximum angle of medialization (degrees) D Maximum angle of lateralization (degrees) E Maximal screw depth (mm) F Minimal screw depth (mm) G Overhang of post arch over entry point (mm) H Lateral mass height, posterior (mm) I Lateral mass height, anterior (mm) J Maximum superiorly directed angle (degrees) K Maximum inferiorly directed angle (degrees) 3.9 7.3 33 13 22.5 14.4 11.4 15.6 17.2 19 0 Median 3.6 7.3 33 13 22.7 14.2 11.5 15.6 17.4 18 0 Standard deviation 1.2 1.3 4 6 2.6 2.0 2.4 2.4 2.1 5 3 Range 1.9 8.0 3.9 10.1 25 45 3 27 15.1 29 10.5 18.5 6.9 17 8.1 20 10.6 22 12 34 85

NEUROSURGERY

VOLUME 54 | NUMBER 6 | JUNE 2004 | 1437

WANG

AND

SAMUDRALA

FIGURE 2. Case illustration of lateral mass screws used to supplement an occipitocervical fusion where the posterior elements of C2 were unsuitable for fixation. Computed tomographic images (A, axial; B, lateral) showing bicortical purchase.

ized trajectory, and minimum screw depth was determined on the basis of a laterally placed screw without any deviation. The mean maximal screw length to obtain bicortical purchase was 22.5 mm, and the mean minimum depth was 14.4 mm. The overhang of the posterior arch averaged 11.4 mm (range, 6.917 mm). This measurement would be the additional screw length needed to keep the polyaxial head of the screw posterior enough to avoid the spinal canal given an idealized trajectory directed 10 to 15 degrees medially. Lateral mass heights varied substantially, but all specimens could easily accommodate a 4-mm screw. This resulted in a large range for screw angulation in the sagittal plane. The mean maximal superior screw angulation was 19 degrees cephalad off a true axial plane. Minimal or no inferiorly directed trajectories were tolerated.

DISCUSSION
Atlantal lateral mass screws represent a significant advance in spinal surgery. The ability to rigidly fixate the C1 vertebra independently of the occiput and axis now provides surgeons

greater latitude in posterior fixation techniques. The initial advantage of C1 lateral mass screws was that they allowed atlantoaxial screw fixation in patients with an atrophic pars. However, it soon became recognized that a strong posterior anchor at C1 could help surgeons avoid unnecessary fusions cranial to the occiput and limit the caudal extensions of occipitocervical instrumentation. Because of the tremendous usefulness of this procedure, it has been disseminated widely since its initial description. However, few anatomic or biomechanical studies have been published on this technique. In two previous anatomic studies, the vast majority of atlas specimens were found to be suitable for lateral mass screw placement (6, 9). In the report by Lynch et al. (9), 98.75% of the 120 specimens examined were able to accommodate 3.5-mm screws. Gupta and Goel (6) confirmed these findings in their examination of 50 bony skeletons, showing that on the basis of morphometric data, all specimens would permit lateral mass fixation. Furthermore, the proximity of the vertebral artery foramina to the lateral atlantoaxial facet joint made vascular injury possible in up to 15% of patients by the standard Magerl screw technique, suggesting that combining C1 lateral mass and C2 pedicle screws may be safer than C1/C2 transarticular screws. Their anatomic study was supported by a series of 106 clinical cases using C1 lateral mass screws, two of which were complicated by asymptomatic vertebral artery violation. This study builds on those initial findings, elaborating in greater detail the anatomy relevant to surgical fixation. Large variations in all of the dimensions of the atlas were found. However, in all specimens, the bony anatomy allowed placement of 3.5-mm screws. In the original description of the technique by Harms and Melcher (7), the screw trajectory was determined by aiming for the anterior arch of C1 under fluoroscopic control, and no particular trajectory angles were specified. Subsequent reports have noted that 10 degrees of medial angulation is ideal, with the sagittal orientation determined by fluoroscopy (17). In this study, the large bulk of the lateral masses was found to permit substantial variations in mediolateral and craniocaudal screw angulation. However, the screw entry points were less forgiving. In 65% of the specimens, the height of the entry point was less than 4 mm, necessitating drilling and removal of the inferior insertion of the dorsal arch. Failure to create an acceptable posterior surface of the lateral mass could easily result in violation of the C1/C2 facet joint. Although this may be of no consequence in cases involving fusion of this motion segment, drill entry through the facet joint itself may result in a screw path predominantly within the synovial space of the joint. The need for careful attention to the screw entry site is also borne out in clinical practice, where the proximity of the vertebral venous plexus, vertebral artery, and C2 root can obscure proper visualization of the entry site. The anatomic study by Doherty and Heggeness (1) reported that the thickest and most dense cortical bone is found in the anterior cortex of the anterior ring. Substantial variations in the density of trabecular bone within the lateral masses were also found, indicating that the strongest screws would be

1438 | VOLUME 54 | NUMBER 6 | JUNE 2004

www.neurosurgery-online.com

ATLANTAL LATERAL MASS SCREW ANATOMY

placed bicortically through the anterior cortex. The variations in vertebral size found in this study emphasize the need for careful intraoperative assessment of screw depth if bicortical purchase is desired. Preoperatively, computed tomographic scans can be used to measure the ideal screw length. Intraoperatively, probe palpation of the anterior cortex and lateral fluoroscopic images can be used to guide screw depths

17. Stokes JK, Villavicencio AT, Liu PC, Bray RS, Johnson JP: Posterior atlantoaxial stabilization: A new alternative to C12 transarticular screws. Neurosurg Focus 12:Article 1, 2002.

Acknowledgments
We thank Bernard Slavin, Ph.D., and the Willed Body Program at the Keck School of Medicine, University of Southern California, for their assistance in this study. This study was supported in part by a restricted research grant from Depuy Acromed, Inc., Raynham, MA.

CONCLUSION
Atlantal lateral mass screws provide the surgeon with an expanded armamentarium for spinal fixation in the high cervical region. The large size of the lateral masses, even in small patients, makes this a relatively forgiving technique. However, preparation of the entry point is vital in most cases to avoid violating the C1/C2 facet joint. The results of this study should aid surgeons in safe and accurate screw placement.

COMMENTS
ang and Samudrala carefully examined 74 cadaveric spines to define the parameters of screw placement into the atlantal lateral masses. This detailed investigation confirms the relative safety of C1 screw fixation. All specimens could accommodate 3.5-mm-diameter screws, and 97% could accept 4-mm-diameter screws. A medial trajectory allows for more bone purchase, and this study shows that a medial trajectory of 25 degrees was possible in all specimens. The most significant anatomic limitation was the screw entry point. In 65% of the specimens, removal of at least a portion of the insertion of the dorsal arch of C1 was necessary to place 4-mm-diameter screws. This undoubtedly decreases the structural integrity of the posterior arch of C1, a situation that may be important if the surgeon is planning to augment the fixation with an atlantal-axial tension band wiring. In such cases, it may be preferable to use a 3.5-mm screw. Vincent C. Traynelis Iowa City, Iowa ang and Samudrala performed anatomic analysis of the C1 lateral mass of 74 cadaveric spines for determining optimal screw placement. Their Table 1 summarizes their findings. They found that maximal screw purchase was achieved with a superior and medial trajectory. The mean maximal screw length required for bicortical purchase using a medial angulation was 22.5 mm, with a range from 15 to 29 mm. The maximal angle of medialization ranged from 25 to 45 degrees, whereas maximum superior angulation ranged from 12 to 34 degrees. From their cadaveric analysis of C1 lateral mass anatomy for screw placement, the authors did not mention their preferred trajectory. Can we infer from their analysis that a safe C1 lateral mass screw placement using the Harms technique would be a medial angulation less than 25 degrees (10 degrees was mentioned as ideal in previous studies) and a superior angulation of approximately 10 degrees? Furthermore, it would be of interest to know whether there were any significant morphometric differences of the C1 lateral mass between male and female specimens or even the left side and the right. In 65% of their cadaveric specimens, drilling and removal of the inferior insertion of the dorsal arch was necessary to accommodate a 4-mm screw. We have also observed this in our own experience of C1C2 screw fixation. In half of our patients, removal of part of the C1 posterior arch facili-

REFERENCES
1. Doherty BJ, Heggeness MH: The quantitative anatomy of the atlas. Spine 19:24972500, 1994. 2. Goel A, Laheri V: Plate and screw fixation for atlanto-axial subluxation. Acta Neurochir (Wien) 129:4753, 1994. 3. Goel A, Desai KI, Muzumdar DP: Atlantoaxial fixation using plate and screw method: A report of 160 treated patients. Neurosurgery 51:13511356, 2002. 4. Gonzalez L, Crawford N, Chamberlain R, Garza LP, Preul M, Sonntag VKH, Dickman C: Craniovertebral junction fixation with transarticular screws: Biomechanical analysis of a novel technique. J Neurosurg 98[Suppl 2]:202 209, 2003. 5. Grob D, Crisco JJ III, Panjabi M, Wang P, Dvorak J: Biomechanical evaluation of four different posterior atlantoaxial fixation techniques. Spine 17: 480490, 1992. 6. Gupta S, Goel A: Quantitative anatomy of the lateral masses of the atlas and axis vertebrae. Neurol India 48:120125, 2000. 7. Harms J, Melcher RP: Posterior C1C2 fusion with polyaxial screw and rod fixation. Spine 26:24672471, 2001. 8. Henriques T, Cunningham BW, Olerud C, Shimamoto N, Lee GA, Larsson S, McAfee PA: Biomechanical comparison of five different atlantoaxial posterior fixation techniques. Spine 25:28772883, 2000. 9. Lynch J, Christensen D, Currier B: C1 lateral mass screws: Technique and morphometric study. Presented at the American Association of Neurological Surgeons meeting, Toronto, ON, Canada, 2001. 10. Melcher RP, Puttlitz CM, Kleinstueck FS, Lotz JC, Harms J, Bradford DS: Biomechanical testing of posterior atlantoaxial fixation techniques. Spine 27:24352440, 2002. 11. Naderi S, Crawford NR, Song GS, Sonntag VKH, Dickman CA: Biomechanical comparison of C1C2 posterior fixations: Cable, graft, and screw combinations. Spine 23:19461956, 1998. 12. Oda I, Abumi K, Sell LC, Haggerty CJ, Cunningham BW, McAfee PC: Biomechanical evaluation of five different occipito-atlanto-axial fixation techniques. Spine 24:23772382, 1999. 13. Paramore CG, Dickman CA, Sonntag VKH: The anatomic suitability of the C12 complex for transarticular screw fixation. J Neurosurg 85:221224, 1996. 14. Resnick DK, Benzel EC: C1C2 pedicle screw fixation with rigid cantilever beam construct: Case report and technical note. Neurosurgery 50:426428, 2002. 15. Resnick DK, Lapsiwala S, Trost GR: Anatomic suitability of the C1C2 complex for pedicle screw fixation. Spine 27:14941498, 2002. 16. Richter M, Schmidt R, Claes L, Puhl W, Wilke HJ: Posterior atlantoaxial fixation: Biomechanical in vitro comparison of six different techniques. Spine 27:17241732, 2002.

NEUROSURGERY

VOLUME 54 | NUMBER 6 | JUNE 2004 | 1439

WANG

AND

SAMUDRALA

tated C1 lateral mass screw placement and prevented fracturing of the arch during placement. Overall, Wang and Samudrala have provided anatomic data relevant to placing C1 lateral mass screws. This will greatly aid many surgeons attempting to place C1 lateral mass screws. Hoang N. Le Daniel Kim Stanford, California ang and Samudrala have provided an extensive analysis of 74 cadaveric spines for C1 lateral mass screw placement. They have described the length of the screws, trajectories, and the size of the screws. They observed a significant radiation and morphology but also note that the large size of the lateral mass makes screw placement forgiving. This study by Wang and Samudrala further substantiates the importance of the C1 lateral mass as a fixation point for occipital cervical surgery. They point out the potential benefits of using C1 as a fixation point but not necessarily obligating fixation to the occiput and fusion to the occiput in selected cases. This is a very meticulous work. Edward C. Benzel Cleveland, Ohio

ositioning the lateral mass screw in C1 coupled with pedicle-screw fixation of C2 is becoming a familiar technique for fusing C1 and C2. The anatomic landmarks of C1 are important. The authors measured these landmarks in 74 cadaveric specimens and concluded that all specimens can accommodate 3.5-mm lateral mass screws bilaterally and that 97% can accommodate 4-mm screws. They also found significant variations in the morphology of C1. Consequently, preoperative computed tomographic scans must be studied carefully. Intraoperative fluoroscopy is also of benefit. As the authors rightly emphasized, the most important aspect of C1 lateral mass screws is the entry site of the screw. We have used these techniques and agree with the authors that the morphology of the C1 lateral mass and the medial angle needed for the screw are quite variable. As the authors have demonstrated, this angle ranges from 25 to 45 degrees medially. The lateral mass of C1 is ovoid and positioned diagonally to the arch of C1. This position can vary quite a bit. The authors have provided us with the detailed measurements of the lateral mass of C1. Volker K.H. Sonntag Phoenix, Arizona

1440 | VOLUME 54 | NUMBER 6 | JUNE 2004

www.neurosurgery-online.com