James T. Guille and Reginald S. Fayssoux
DEFINITION
Reduction of scoliotic spinal deformity via posterior spinal fusion with instrumentation allows for improvement of the cosmetic appearance of the child or adolescent with scoliosis and prevents curve progression. The goal is to balance these advantages with the inherent risks of instrumentation and reduction maneuvers.
Instrumentation provides an internal construct holding the spine in its corrected position until spinal fusion is achieved (about 6 months) and obviates the need for postoperative immobilization. A Cobb angle measurement greater than 10 degrees distinguishes minor spine asymmetry from true scoliosis.
Segmental instrumentation with hooks and pedicle screws provides multiple fixation points, allowing for three dimensional correction of the scoliotic spine.
Instrumentation is introduced after posterior exposure of the thoracic or lumbar spine (see Chap. PE-56).
ANATOMY
A thorough knowledge of scoliotic spinal anatomy is critical for the safe placement of instrumentation.
In the scoliotic spine there is rotation of the vertebral bodies in the transverse plane, with the spinous processes rotating toward the concavity of the curve.
Anatomy of the posterior elements
In the scoliotic spine, the pedicles on the concave side are narrower than those on the convexity, especially at the apex of the curve (FIG 1A).12
The intervertebral foramina in the thoracic spine are larger and deeper than in any other region of the spine, with the exiting nerve root occupying less than 25% of the foramen and coursing through its midportion.
Lumbar nerve roots pass adjacent to the inferomedial aspect of the pedicle and lie superior within the foramina.
Scoliotic deformity affects not only the bony anatomy but also the relationship of the spine to the adjacent soft tissue elements.
The aorta is posterolateral to its normal position, putting it at risk for injury with left-sided lateral pedicular breaches (FIG 1B).
The spinal cord hugs the concavity of the curve such that the width of the epidural space is less than 1 mm at the thoracic apical vertebral levels on the concave side; it is 3 to 5 mm on the convex side.8
Thoracic Spine Anatomy
The thoracic facets are more coronally oriented in comparison to the more sagittally oriented lumbar facets.
Pedicle anatomy for placement of pedicle screws
Dimensions
In scoliotic spines, average thoracic pedicle length (distance from the posterior cortical starting point to the posterior longitudinal ligament in line with the axis of the pedicle) is 16 to 22 mm.
Average thoracic cord length (distance from the posterior cortical starting point to the anterior vertebral cortex in line with the axis of the pedicle) is 34 to 52 mm; it is typically greater on the concavity.
Coronal anatomy
In the scoliotic spine, the medial wall is two to three times thicker than the lateral wall at all thoracic levels. This may be why most screw-related pedicle fractures occur laterally.
FIG 1 • A. Comparison of a normal thoracic vertebra on the left and a scoliotic thoracic vertebra on the right. Note the narrowed concave pedicle on the scoliotic vertebra. B. Thoracic-level axial magnetic resonance imaging in a patient with a right thoracic scoliotic curve. Note the posterolateral position of the aorta. (B: From Sucato DJ, Duchene C. The position of the aorta relative to the spine: a comparison of patients with and without idiopathic scoliosis. J Bone Joint Surg Am 2003;85A:1461–1469. Reprinted with permission from The Journal of Bone and Joint Surgery, Inc.)
The coronal anatomy of the thoracic pedicle varies, moving from anterior to posterior. The likelihood of pedicle wall breach is greatest midway between the lamina and body with placement of screws.
Pedicle width decreases from T1 to T4 and then gradually increases to T12, while pedicle height and length tend to increase from T1 to T12.
Main pedicle diameter is 4 to 6 mm in thoracic scoliosis.
Endosteal pedicle width in the apical region of the thoracic spine measures 2.5 to 4.2 mm on the concavity of the curve and 4.1 to 5.0 mm on the convexity of the curve.
In the thoracic spine, transverse processes do not align with the pedicle in the axial plane: they are rostral to the pedicle in the upper thoracic spine and caudal to the pedicle in the lower thoracic spine (crossover occurs at T6–7).
Transverse orientatio.
T12 pedicles are perpendicular to the floor in the transverse plane.
T1 pedicles subtend an angle of about 25 to 30 degrees with the midline in the transverse plane.
Thoracic pedicles progressively angle outward in the transverse plane, proceeding superiorly from T12 to T1.
Thoracic pedicle screw starting point.
As one proceeds proximally from T12 there is a trend toward a more medial and cephalad pedicle starting point as one proceeds to the apex of the thoracic spine (FIG 2 and Table 1).
This then transitions to a trend toward a more lateral and caudal pedicle starting point as one proceeds proximally from the apex.
FIG 2 • Thoracic vertebrae starting points (see Table 1).
Lumbar Spine Anatomy
The lumbar vertebral facets are more sagittally oriented in comparison to thoracic vertebral facets.
Pedicles
Dimensions
In scoliotic spines, average lumbar pedicle length is 20 to 22 mm.
Average lumbar cord length is 45 to 48 mm.
Coronal anatomy
Average lumbar endosteal pedicle width is 4.8 to 9.5 mm.
The larger size of the lumbar pedicles increases the likelihood of optimal placement of pedicle screws.
Transverse orientation
L1 pedicles are perpendicular to the floor in the transverse plane.
L5 pedicles subtend an angle of about 25 to 30 degrees with the midline in the transverse plane.
Lumbar pedicles progressively angle outward in the transverse plane, proceeding inferiorly from L1 to L5.
Lumbar pedicle screw starting points
The long axis of the pedicle pierces the lamina at the intersection of two lines: a vertical line tangential to the lateral border of the superior articular process, and a horizontal line bisecting the transverse process (FIG 3).
The point of intersection for these two lines lies in the angle between the superior articular process and the base of the transverse process.
Dangers
Medial pedicular breaches endanger the dural sac, especially on the concavity of the curve.
Inferior pedicular breaches endanger the nerve root, especially in the lumbar spine.
Advancement of pedicle screws following a lateral pedicular breach on the left can endanger the lung, segmental vessels, and sympathetic chain (T4–T12) and the aorta (T5–T10).
FIG 3 • Lumbar vertebrae starting points.
Advancement of pedicle screws following a lateral pedicular breach on the right can endanger the lung, segmental vessels, sympathetic chain, and azygous vein (T5–T11).
Advancement of pedicle screws following a breach of the anterior cortex on the right can endanger the superior intercostal vessels (T4–T5), the esophagus (T4–T9), the azygous vein (T5–T11), the inferior vena cava (T11–T12), and the thoracic duct (T4–T12).
Advancement of pedicle screws following a breach of the anterior cortex on the left can endanger the esophagus (T4–T9) and the aorta (T5–T12).
PATHOGENESIS, NATURAL HISTORY, AND PATIENT HISTORY AND PHYSICAL FINDINGS
See Chapter PE-56, Posterior Exposure of the Thoracic and Lumbar Spine.
IMAGING AND OTHER DIAGNOSTIC STUDIES
Placement of a pedicle screw at the thoracolumbar junction followed by intraoperative fluoroscopic imaging accurately identifies vertebral level.
With use of intraoperative fluoroscopic imaging guidance, knowledge of anatomy remains critical in order to orient the intensifier to obtain the best coronal images of the pedicles.
Radiographic criteria used to evaluate accurate screw placement
Harmonious segmental change of the tips of the pedicle screws on the posteroanterior (PA) radiograph
No crossing of the medial pedicle wall by the tip of the pedicle screw with reference to vertebral rotation on the PA radiograph
No violation of the imaginary midline of the vertebral body by the tip of the pedicle screw on the PA radiograph
No breach of the anterior cortex of the vertebral body on the lateral radiograph
DIFFERENTIAL DIAGNOSIS
Scoliosis
Idiopathic
Congenital
Neuromuscular
Limb-length discrepancy
Osteoid osteoma
Sprengel deformity
NONOPERATIVE MANAGEMENT
Observation for progression for curves of 0 to 20 degrees. Patients are followed with serial clinical and radiographic examinations.
Bracing for progressive curves of 20 to 40 degrees if the patient is skeletally immature. Braces are unable to correct curves; their purpose is to prevent curve progression.
SURGICAL MANAGEMENT
Preoperative Planning
PA and lateral radiographic views of the entire spine.
Supine bending radiographs may show pedicles not visible on the PA view.
Convex apical pedicles greater than 5 mm in diameter on the PA radiographs are large enough to accommodate pedicle screw placement.
Computed tomographic (CT) imaging can be used to evaluate pedicle morphology, with images oriented perpendicular to the plane of the vertebrae.
Fusion levels are chosen based on the Lenke criteria.
Hook Placement
Advantages
Technically easier, especially at levels with small pedicles (apical concave pedicles)
Less operative time
Disadvantages
Increased canal intrusion in comparison to pedicular fixation
Lack of three-column fixation
Decreased ability to perform correctional derotational maneuvers
Types of hooks
Laminar hooks: should be used with caution in the neurologically intact patient
Thoracic laminar hooks: downgoing sloped hooks
Lumbar laminar hooks: upgoing or downgoing Cshaped hooks
Thoracic transverse process hooks: downgoing C-shaped hooks
Thoracic pedicle hooks: upgoing claw-tip hooks placed under the lamina
Offset hooks: offset laterally to lie in line with pedicle screws for use in hybrid constructs
Hook patterns for an isolated thoracic curve
Concave side
Upgoing pedicle hook on upper end and upper intermediate vertebrae
Downgoing laminar hook on lower end and lower intermediate vertebra
Convex side
Claw construct at upper end vertebrae.
Downgoing transverse process hook with upgoing pedicle hook at the same level or next-distal level. Splitting the claw over two levels (split claw) better resists rotational forces.
Pedicle Screw Placement
Advantages
Pedicle screws have significantly higher axial pullout strengths than supralaminar hooks and pedicle hooks.
Better correction in the coronal and axial planes.
Less decrease in pulmonary function than anterior surgery.
No implants in the canal during the correction phase.
Correction not gained by pure distraction.
Fewer fusion levels.
No crankshaft.
Larger area for bone graft.
Allows earlier postoperative activity.
Disadvantages
Steep learning curve.
Caudal or medial penetration can result in dural or neural injury.
Lateral penetration can cause vascular injury.
Increased operative time.
Costly procedures.
Complications
Suboptimal screw position
More common in cases of severe deformity
Perforation not uncommon (up to 40% of screws in some series)
Lateral perforation more common than medial perforation
Lowest containment rates in midthoracic spine (T5 to T8)
Dural, neural, or vascular injuries occur infrequently.
Types of pedicle screws
Monoaxial
No motion between the screw and the screw head
Can obtain axial correction of deformity
Uniaxial
Motion between the screw and the screw head constrained to one plane
Can accommodate sagittal contours while retaining ability to obtain axial correction (derotation)
Polyaxial
Multiaxial motion allowed between screw and screw head
For accommodation of sagittal contours
Can accommodate malalignment of the starting points in the coronal plane
Reduction screw
Pedicle screw with breakaway extended tabs
Useful for seating rod into pedicle screw for difficult reduction maneuvers
Freehand placement of thoracic pedicle screws
The straightforward trajectory allows for fixed-head screws and true direct vertebral derotation.
Anatomic trajectory has a longer bone channel and allows a longer screw to be placed, but mandates the use of a multiaxial screw to connect it to the rod.
A straightforward trajectory paralleling the superior endplate has significantly higher pullout strength versus an anatomic trajectory that angles about 22 degrees in the cephalocaudal direction perpendicular to the superior facet.7
Extrapedicular thoracic pedicle screws
Screw inserted at the junction of the cephalad tip of the transverse process and the rib with advancement caudad so that the screw is contained in the pedicle rib unit, defined as the space between the lateral pedicle cortex and medial rib cortex.
Similar biomechanical fixation strength of transpedicular screws.
Confirmation of screw placement
Radiographic confirmation (see Imaging and Other Diagnostic Studies)
Neuromonitoring
Electromyography to confirm intraosseous screw placement
Somatosensory evoked potentials and motor evoked potentials
Postoperative CT scan routinely performed before patient leaves the hospital to confirm accurate screw placement
Positioning
Patient is intubated in the supine position on the stretcher.
Neurologic monitoring leads are placed cranially, on the intercostal and abdominal musculature, and on all four extremities.
Multiple large-bore intravenous access is obtained for fluid management and an arterial line is placed for intraoperative blood pressure monitoring.
The patient is transferred to the prone position on a well-padded operating room table such as a Jackson frame (Orthopaedic Systems, Union City, CA).
Care should be given to the degree of hip flexion– extension, as this can affect the amount of lordosis in the lumbar spine.
Bolsters underneath the chest and anterior superior iliac spines prevent abdominal compression and allow epidural venous return, thus decreasing epidural bleeding.
All bony prominences are well padded, including medial elbows, knees, pretibial areas, and ankles.
Care is taken to avoid abduction and forward flexion past 90 degrees at the shoulder and flexion past 90 degrees at the elbow.
Skin is shaved if excessively hairy.
Clear adhesive surgical drapes (3M Steri-Drape Towel Drapes) are placed around the perimeter of the surgical site, extending from the hairline to the top of the gluteal crease (regardless of levels to be fused, the entire spine should be draped).
If a wake-up test is going to be used by the surgical team, a clear plastic C-arm cover or equivalent clear drape is laid over the exposed feet for visualization during the test.
A disposable plastic ruler used for measuring the pedicle probe for pedicle depth is placed caudal to the field on the buttocks and covered with a clear Tegaderm dressing.
TECHNIQUES
For posterior exposure of the thoracic and lumbar spine for placement of instrumentation, refer to Chapter PE-56.
HOOK PLACEMENT
Proper hook-site preparation is critical to obtain a stable construct and minimize the chance of hardware failure.
Ideally, hooks should be placed flush with the bony surfaces to evenly distribute forces and minimize the chance of hook pullout. This is accomplished by meticulous removal of the soft tissues and judicious contouring of the bony surfaces: removing too much bone can weaken hook purchase, whereas removing too little bone can result in improper seating of the hook.
Pedicle Hooks
Initial pedicle hook-site preparation requires removal of a small portion of the inferior facet with an osteotome (TECH FIG 1A,B). The inferior facet of the superior vertebra is osteotomized using an osteotome. A vertical cut is made at the medial edge of the facet, near the base of the spinous process. A horizontal cut in the inferior facet, allowing removal of 3 to 4 mm of bone, follows for insertion of the pedicle hook.
The exposed hyaline cartilage from the facet joint is removed. A pedicle finder is introduced into the facet joint and gently impacted into place with a mallet (TECH FIG 1C,D). Care is taken to avoid canal penetration.
The permanent pedicle hook is subsequently inserted (TECH FIG 1E,F). The surgeon must carefully visualize the facet joint and avoid inserting the pedicle hook intothe laminae. If the lamina is split by the hook, fixation will be compromised.
Laminar Hooks
Laminar hooks are placed in a similar fashion but require great care because they are positioned in the spinal canal (TECH FIG 2A,B).
To obtain entrance into the canal, the ligamentum flavum is carefully dissected from the laminae and completely removed with curettes and rongeurs until the dura can be visualized. Small laminotomies are performed to allow room for hook insertion (TECH FIG 2C,D).
A hook starter is then used to create a path for the hook (TECH FIG 2E). The hook is then placed into the path created (TECH FIG 2F).
Transverse Process Hooks
Transverse process hooks do not require extensive bone removal but may require minimal contouring to optimize fit (TECH FIG 3A,B).
The costotransverse ligaments on the superior side of the transverse process are divided with a periosteal elevator. The transverse process hook is seated around the transverse process (TECH FIG 3C,D).
TECH FIG 1 • Placement of pedicle hook. A,B. A small portion of the inferior facet is osteotomized. C,D. Introduction of pedicle finder into facet joint, taking care to avoid canal penetration. E,F. Insertion of pedicle hook.
TECH FIG 2 • Placement of supralaminar hook. A,B. Placement of a supralaminar hook is difficult without bone removal to allow room for hook insertion. C,D. Placement of supralaminar hook. Laminotomies allow room for hook insertion. E. Path for hook created with hook starter. F. Insertion of supralaminar hook.
TECH FIG 3 • A,B. Placement of transverse process hook. Contouring of the transverse process is required to optimize fit. C,D. Placement of transverse process hook. After contouring the transverse process, the transverse process hook is seated.
EXPOSURE FOR THORACIC PEDICLE SCREW PLACEMENT
Full exposure of the facet joint, the pars interarticularis, and the entire transverse process aids in identification of the ideal starting point.
Once the entire spine is exposed, each level is considered independently. The surgeon needs to visualize the local topical anatomy and the effects of the scoliosis on the anatomy (rotation).
Bovie cautery is used to outline the osseous anatomy at each level. This first entails finding the medial and lateralborder of the facet and the superior and inferior borders of the transverse process (TECH FIG 4A).
The inferior facet of the superior vertebrae is osteotomized using an osteotome at the medial edge of the facet (TECH FIG 4B) and at the inferior border of the superior transverse process (TECH FIG 4C).
At this point, the facet joint should be fully exposed (TECH FIG 4D), thus facilitating identification of the starting point.
TECH FIG 4 • A. The medial and lateral borders of the facet joint are identified. B. The inferior facet of the superior vertebrae is osteotomized at its medial edge. C. The facet is osteotomized at the inferior border of the superior transverse process. D. Full exposure of the facet joint facilitates identification of the starting point.
THORACIC PEDICLE SCREW PLACEMENT
A 3.5-mm acorn-tipped burr is then used to create a hole at the desired starting point (see Thoracic Anatomy) (TECH FIG 5A). A cancellous blush often heralds entry into the pedicle but can be a false positive found on entry into the transverse process.
A specialized thoracic probe with a 2-mm blunt tip and a 35-mm curved segment with a rectangular cross section (Lenke probe) is used to create the tract for the pedicle screw.
The probe is introduced into the starting point with the curvature oriented so the tip is pointed laterally to avoid medial pedicle cortical violation (TECH FIG 5B).
The cancellous soft spot signifies entry into the pedicle.
The probe is advanced using ventral pressure and axial rotation to a depth of about 15 to 20 mm (the length of the pedicle), using the appropriate orientation for the particular vertebral level (see Thoracic Pedicle Anatomy) and taking care to account for the scoliotic deformity.
The probe is then removed and reintroduced into the previously developed tract with the tip turned medial to avoid lateral vertebral body cortical violation (TECH FIG 5C).
The probe is advanced to a depth appropriate for the particular vertebral level, taking care to avoid anterior and lateral cortical violation.
Typical cord lengths (distance from posterior cortical starting point to anterior vertebral cortex in line with the axis of the pedicle):
Lower thoracic 40 to 45 mm
Midthoracic 35 to 40 mm
Upper thoracic 30 to 35 mm
The tract is probed using a flexible sound and five distinct bony borders are palpated: superior, inferior, medial, and lateral walls and the floor.
The first 15 to 20 mm of the tract corresponds to the pedicle; its integrity should be critically assessed.
The depth is measured with the flexible sound in the base of the tract using a hemostat (TECH FIG 5D).
The tract is tapped (TECH FIG 5E). Undertapping by 1.0 mm creates a 93% increase in maximal screw insertion torque.5
The tract is again probed for a breach with the aid of the feel of the tapped threads.
Screw diameter is based on radiographic evaluation of pedicles.
The screw is placed slowly to allow for viscoelastic expansion of the pedicle (TECH FIG 5F).
TECH FIG 5 • A. A burr is placed at the correct starting point. B. After the cortex is breached, a curved probe is placed into the pedicle with the tip pointing laterally to minimize risk of medial pedicle breach and potential cord injury. C. After probing past the pedicle, its tip is then turned to point medially to minimize risk of vertebral body cortical breach. D. After probing the tract, the depth is measured. E. The tract is tapped. F. The pedicle screw is inserted.
LUMBAR PEDICLE SCREW PLACEMENT
Full exposure of the facet joint, the pars interarticularis, and a portion of the transverse process aids in identification of the ideal starting point.
The facet joint is removed to obtain a flat surface before placing the pedicle screw.
A 3.5-mm acorn-tipped burr is then used to create a cortical breach at the desired starting point.
A specialized probe with a blunt spatula tip and a 35-mm curved segment with a rectangular cross section (lumbar probe) is used to create the tract for the pedicle screw.
The probe is introduced into the starting point with the curvature oriented so the tip is pointed laterally to avoid medial pedicle cortical violation.
The cancellous soft spot signifies entry into the pedicle.
The probe is advanced using ventral pressure and axial rotation to a depth of about 15 to 20 mm (the length of the pedicle), using the appropriate orientation for the particular vertebral level (see Lumbar Pedicle Anatomy) and taking care to account for the scoliotic deformity.
The probe is then removed and reintroduced into the previously developed tract with the tip turned medial to avoid lateral vertebral body cortical violation.
The probe is advanced to a depth appropriate for the particular vertebral level, taking care to avoid anterior and lateral cortical violation.
The tract is probed using a flexible sound and five distinct bony borders are palpated: superior, inferior, medial, and lateral walls and the floor.
The first 15 to 20 mm of the tract corresponds to the pedicle; its integrity should be critically assessed.
The depth is measured with the flexible sound in the base of the tract using a hemostat.
The tract is tapped. Undertapping by 1.0 mm creates a 93% increase in maximal screw insertion torque.5 The tract is again probed for a breach with the aid of the feel of the tapped threads.
Screw diameter is based on radiographic evaluation of pedicles.
The screw is placed slowly to allow for viscoelastic expansion of the pedicle.
ROD PLACEMENT
Stainless steel is preferred over titanium in patients in whom a derotation maneuver is planned because of the higher modulus of elasticity. The disadvantage is failure occurring at the screw–bone interface instead of via plastic deformation of the rod.
Titanium instrumentation allows for magnetic resonance imaging with less metallic artifact and is indicated in patients with nickel allergies or if there are plans to follow the patient with serial MRIs.
We use the Bovie cord to measure the length of rod required, although any flexible measuring template may be used (TECH FIG 6).
TECH FIG 6 • The length of rod required is determined using the Bovie cord.
Typically, 1 cm is added to the measurement of the concave side to allow for distraction. As most corrective maneuvers are done with the concave rod, the length of the convex rod typically mimics the length of the Bovie cord.
The rod is prebent to the appropriate sagittal contour using French benders.
The concave-side rod is placed first.
There are various methods for rod placement based on the corrective measures that are to be used. We prefer to secure the rod to the hook–screw proximally and work distally.
A 90-degree rod derotation maneuver is performed with vise grips. All set screws are then secured to hold correction.
The prebent convex-side rod is placed in situ.
Compression and distraction maneuvers are performed where needed. In general, care should be taken to “horizontalize” the end vertebrae.
For selective fusion of Lenke 1A curves, we prefer to horizontalize the lowest instrumented vertebrae (LIV).
For Lenke 1B curves we prefer to leave a slight obliquity to the LIV.
For Lenke 1C curves it may be preferable to leave the
LIV oblique to prevent coronal decompensation.
Derotation maneuvers can be done to address apical rotation.
Set screws are tightened to appropriate torque using torque wrench.
POSTOPERATIVE CARE
No postoperative immobilization is required with multisegmental constructs.
Postoperative restrictions include limitations with lifting, bending, and twisting.
It is important to maintain mean arterial blood pressure above 70 mm Hg overnight and hemoglobin above 10 g/dL to maintain spinal cord perfusion.
Intravenous antibiotics are maintained for 48 hours postoperatively.
Neurovascular checks are made every 2 hours for the first 8 hours and then every 8 hours.
Patients are out of bed on postoperative day 1.
Foley catheter is removed on postoperative day 2.
Diet is advanced as tolerated.
Patient-controlled analgesia is used for appropriate patients. Continuous narcotic infusion with demand for the first 24 hours is followed by demand only for the next 24 hours, followed by oral pain medications when tolerating diet.
A 4-day hospital course is typical.
Routine follow-up is done at 1, 3, and 6 months and at 1, 2, and 5 years.
Activity is increased based on the degree of fusion.
OUTCOMES
With meticulous attention to detail with regard to instrumentation and fusion techniques, excellent outcomes in terms of straightening and fusion of the scoliotic spine can be expected.
Long-term outcomes are variable and depend on the underlying diagnosis and the extent of retained spinal mobility.
COMPLICATIONS
Wrong-level surgery
Failure of fusion
Hardware malfunction
Neurologic injury
Dural tear
Pneumothorax
Crankshaft
Superior mesenteric artery syndrome
Loss of lumbar lordosis
REFERENCES
· Karim A, Mukherjee D, Gonzalez-Cruz J, et al. Accuracy and safety of thoracic pedicle screw placement in spinal deformities. J Spinal Disord Tech 2005;18:522–526.
· Kim YJ, Lenke LG, Bridwell KH, et al. Free hand pedicle screw placement in the thoracic spine: is it safe? Spine 2004;29:333–342.
· Kim YJ, Lenke LG, Bridwell KH. Comparative analysis of pedicle screw versus hook instrumentation in posterior spinal fusion of AIS. Presented at SRS, Quebec City, September 2003.
· Kim YJ, Lenke LG, Cheh G, et al. Evaluation of pedicle screw placement in the deformed spine using intraoperative plain radiographs: a comparison with computerized tomography. Spine 2005; 30:2084–2088.
· Kuklo TR, Lehman RA Jr. Effect of various tapping diameters on insertion of thoracic pedicle screws: a biomechanical analysis. Spine 2003;28:2066–2077.
· Kuklo TR, Lenke LG, O'Brien MF, et al. Accuracy and efficacy of thoracic pedicle screws in curves more than 90 degrees. Spine 2005; 30:222–226.
· Lehman RA, Polly DW Jr, Kuklo TR, et al. Straight-forward versus anatomic trajectory technique of thoracic pedicle screw fixation: a biomechanical analysis. Spine 2003;28:2058–2065.
· Liljenqvist UR, Allkemper T, Hackenberg L, et al. Analysis of vertebral morphology in idiopathic scoliosis with use of magnetic resonance imaging and multiplanar reconstruction. J Bone Joint Surg Am 2002;84A:359–368.
· Liljenqvist UR, Link TM, Halm HF. Morphometric analysis of thoracic and lumbar vertebrae in idiopathic scoliosis. Spine 2000;25: 1247–1253.
· O'Brien MF, Lenke LG, Mardjetko S, et al. Pedicle morphology in thoracic adolescent idiopathic scoliosis: is pedicle fixation an anatomically viable technique? Spine 2000;25:2285–2293.
· Parent S, Labelle H, Skalli W, et al. Thoracic pedicle morphometry in vertebrae from scoliotic spines. Spine 2004;29:239–248.
· Parent S, Labelle H, Skalli W, et al. Morphometric analysis of anatomic scoliotic specimens. Spine 2002;27:2305–2311.
· Sucato DJ, Duchene C. The position of the aorta relative to the spine: a comparison of patients with and without idiopathic scoliosis. J Bone Joint Surg Am 2003;85A:1461–1469.
· Vaccaro AR, Rizzolo SJ, Allardyce TJ, et al. Placement of pedicle screws in the thoracic spine. Part I: Morphometric analysis of the thoracic vertebrae. J Bone Joint Surg Am 1995;77A:1193–1199.