Spinal Instability and Spinal Fusion Surgery

Updated: Jan 13, 2016
  • Author: Peyman Pakzaban, MD; Chief Editor: Brian H Kopell, MD  more...
  • Print


In the past 3 decades, increased understanding of spinal biomechanics, proliferation of sophisticated spinal instrumentation devices, advances in bone fusion techniques, refinement of anterior approaches to the spine, and development of microsurgical and minimally invasive methods have made it possible to stabilize every segment of the spine successfully, regardless of the offending pathology. Accordingly, use of spinal fusion and instrumentation has increased. The question facing the modern spine surgeon is not so much how to stabilize the spine but when to do so.


History of the Procedure

Spinal fusion and instrumentation were developed and applied as independent techniques for treatment of spinal instability in the first half of the 20th century, before the biomechanical principles surrounding spinal instability were understood.

Around the turn of the 20th century, the problem of progressive spinal deformity and disability caused by spinal tuberculosis (Pott disease) had become a focus of clinical inquiry. The problem did not yield itself to the decompressive procedures (eg, laminectomy) developed in the previous century. In 1911, Russell Hibbs and Fred Albee independently developed the concepts and methods for bony fusion of the spine to address the symptoms of Pott disease. These methods and their subsequent refinements consisted of applying autologous bone (harvested from laminae, iliac crest, or ribs) to the dorsal surface of spine. Although this constituted a major advance in spine surgery that was subsequently applied to a much wider range of pathological disorders and which remains in use today, the method of onlay posterior grafting, when performed in isolation, suffered from an unacceptably high rate of pseudarthrosis (failed fusion).

Around this time, spinal instrumentation, which mostly consisted of wiring of posterior elements, was employed sporadically for treatment of spine fractures. This method was first employed by Berthold Hadra in 1891. In the 1950s, Paul Harrington pursued his historic work on correction of idiopathic and postpolio scoliosis by applying a combination of compression and distraction hooks and rods to the thoracolumbar spine. [1] The success of the Harrington rod system with deformity correction led to its subsequent use for treatment of overt spinal instability (eg, post-traumatic instability). However, it soon became apparent that the application of spinal instrumentation (without fusion) for treatment of spinal instability often ended in breakage or loosening of the hardware (hardware failure).

Harrington later expressed the idea that there is a “race between instrumentation failure and acquisition of spinal fusion.” This principle and the realization that the problems of pseudarthrosis and hardware failure could be resolved if bone grafting and instrumentation were used together laid the foundations of modern spine stabilization surgery. In current practice, bone grafting and instrumentation are often used concurrently based on the expectation that internal fixation of spine enhances the success of bone fusion while a successful bone fusion eliminates the possibility of hardware failure by reducing the chronic biomechanical stresses on the hardware construct.

Of note, the term "fusion" is used in this article and in spine literature to refer to the concept of internal stabilization of spine, generally accomplished by fusion with instrumentation (instrumented fusion), but also, albeit with decreasing frequency, accomplished by bone grafting alone.



Strictly defined, spinal fusion is an operation designed to treat spinal instability. In practice, however, this definition is not particularly useful as it fails to establish the indications for spinal fusion. The problem is threefold: (1) the current definitions of spinal instability are not uniformly accepted and applied; (2) it is difficult to measure instability in individual clinical circumstances; and (3) class I and II scientific evidence regarding spinal fusion is scarce. In this setting, clinical practice is guided by an understanding of the principles of spinal biomechanics and knowledge of the generally accepted indications, contraindications, and controversies regarding fusion surgery.

In their widely-quoted work, White and Panjabi defined spinal stability as the ability of the spine under physiological loads to limit patterns of displacement so as to not damage or irritate the spinal cord and nerve roots and, in addition, to prevent incapacitating deformity or pain due to structural changes. [2] Conversely, instability refers to excessive displacement of the spine that would result in neurological deficit, deformity, or pain. Instability can be acute (eg, spine fractures and dislocations) or chronic (eg, spondylolisthesis). Acute instability has been further subcategorized as overt versus limited, whereas chronic instability has been subdivided to include glacial instability (progressive deformity) and instability associated with dysfunctional motion segment. [3]

A simpler conceptual approach would be to think of instability as overt, anticipated, or covert.

Overt instability refers to excessive motion that is readily documented by radiographic studies and results in pain, deformity, or neurological deficit. Those spine fractures, dislocations, tumors, and infectious processes that significantly disrupt one or more spinal motion segments produce acute overt instability. Spondylolisthesis with abnormal dynamic displacement, documented on flexion/extension x-ray films, is an example of chronic overt instability. In addition, any spinal deformity (kyphosis, hyperlordosis, scoliosis, or spondylolisthesis) that progresses with time as documented by serial radiographs (ie, Benzel glacial instability) falls in the category of chronic overt instability. Overt instability generally requires stabilization, either by external means (bracing) or internal means (fusion). [4]

Bilateral jumped facet syndrome is example of over Bilateral jumped facet syndrome is example of overt spinal instability due to trauma. Note grossly abnormal displacement of C5 relative to C6 with neck flexion.

Anticipated instability refers to instability that would be produced by a surgical procedure that is required for proper decompression of neural elements or resection of an offending lesion. For instance, corpectomy or total facetectomy would constitute indications for fusion at the time of the original operation. A comprehensive anterior cervical discectomy (with complete resection of the posterior longitudinal ligament and portions of both uncovertebral joints performed for adequate neural decompression) may also be considered in this category, as its disrupts 2 of Denis' 3 spinal columns.

Example of anticipated instability. Figure A shows Example of anticipated instability. Figure A shows large mass affecting right C3-4 facet joint and lateral masses in patient with severe right-side neck and shoulder pain; Figures B and C show complete resection of tumor and simultaneous C3-4 anterior fusion to circumvent anticipated iatrogenic stability produced by radical resection of facet and lateral masses.

Covert instability is a more elusive concept. It refers to circumstances in which excessive motion cannot be grossly demonstrated but is presumed to exist based on the combination of clinical and radiographic findings. Fixed spondylolisthesis (without movement on flexion and extension x-ray films) in the setting of progressively worsening back pain and/or radicular symptoms is a good example of covert instability. Pseudarthrosis with intact instrumentation also falls in this category. Controversy arises when the concept of covert instability is applied to degenerative diseases of the spine. In this context, the concept of micro-instability is sometimes evoked to justify fusion for a wider range of conditions, including recurrent disc herniation, disc degeneration with discogenic pain, painful facet arthropathy, spinal stenosis, and failed back syndrome without overt instability.

Spinal stenosis with fixed degenerative spondyloli Spinal stenosis with fixed degenerative spondylolisthesis in elderly patient is common example of covert instability. Acceptable surgical treatment options include decompression alone vs decompression with fusion.



Since spinal instability is not a single disease but a pathological consequence of a variety of different spine disorders such as traumatic fractures, metastatic tumors, [5] and degenerative conditions, each with its own epidemiology, it is not possible or meaningful to determine the incidence and prevalence of spinal instability in the population. Furthermore, because of the disagreements on indications for spine fusion (at least for degenerative disease), the incidence of spinal instability does not correlate with the observed frequency of spine fusion surgery.

More than 400,000 spinal fusions are performed in the United States annually. The vast majority of these operations are performed for degenerative disease of the spine. Between 1998 and 2008, the annual number of spinal fusion discharges increased 137%, from 174,223 to 413,171 (P<0.001), and the mean age for spinal fusion increased from 48.8 to 54.2 years (P<0.001). [6]

Between 1996 and 2001, the number of spine fusions in the United States increased 76%. [7] Whereas in 1990 about 70% of cervical spine operations consisted of nonfused decompressions, by 2000 about 70% of cervical spine operations consisted of anterior cervical fusions. [8] An increase in the incidence of spinal instability could certainly not account for the increase in fusion surgery. While the forces driving this trend are debated, the standard of care in the United States is clearly shifting toward greater use of fusion surgery.



Virtually every category of disease affecting the bones, discs, joints, or ligamentous support structures of the spine can produce spinal instability. These include trauma, tumors, infections, inflammatory diseases, connective tissue disorders, congenital disorders, degenerative disorders, and iatrogenic (postsurgical) etiologies.



The pathophysiology of spinal instability is variable and dependent on the etiology of instability. However, an understanding of certain biomechanical principles can guide the surgeon in diagnosing spinal instability and selecting the appropriate treatment method.

The 3-column concept of the spine as defined by Denis is widely used as the conceptual framework for diagnosing acute overt spinal instability. [9] Although originally devised based on a retrospective review of traumatic injuries to the thoracic and lumbar spine, it is now also applied to the subaxial (below C2) cervical spine and to nontraumatic instability. The anterior column consists of the anterior vertebral body (usually anterior two-thirds), the anterior annulus, and the anterior longitudinal ligament. The middle column refers to the posterior wall of the vertebral body, the posterior annulus, and the posterior longitudinal ligament. The posterior column refers to the posterior ligamentous complex that connects adjacent neural arches, consisting of facet capsules, ligamentum flavum, interspinous ligament, and supraspinous ligament. Failure of two or more columns generally results in instability.

In this context, a simple compression wedge fracture occurs due to failure of the anterior column with preservation of the middle column (stable). On the other hand, a burst fracture occurs due to compression failure of both anterior and middle columns (usually unstable), often resulting in bone retropulsion into spinal canal. A seat-belt type injury is attributed to distraction failure of the posterior and middle columns with hinging of an intact anterior column (unstable). Fracture-dislocations involve failure of all 3 columns and are considered highly unstable.

Instantaneous axis of rotation (IAR) is the axis about which a vertebral segment would rotate when exposed to an asymmetric application of force. Although theoretically there are 3 axes of rotation corresponding to rotation in the sagittal plane (flexion/extension), coronal plane (lateral bending) and axial plane (twisting), most references to IAR correspond to axial forces applied in the sagittal plane. IAR commonly (but not necessarily) falls within Denis' middle column.

Force vectors are simple mathematical constructs that define not only the magnitude of a force, but also its direction. A force vector applied at a distance to the IAR results in rotation of that vertebral segment about the IAR. The distance between the point of application of the force vector and the IAR is called the moment arm. The longer the moment arm, the less force is required to produce rotation.

When unrestricted rotation or displacement of an object is not possible in response to a force vector, deformation of its material (in this case bone) occurs. For solid objects, elastic deformation occurs if the material can resume its shape when the stress (force divided by cross-sectional area) is removed. With increasing stress, a threshold is reached (elastic yield point) beyond which irreversible but smooth deformation (plastic deformation) occurs. With further increases in stress, another threshold is reached (ultimate tensile point or failure point), at which point a fracture occurs and the stress is relieved. In the case of vertebral bone, the elastic yield point and failure point are fairly close, so the very little plastic deformation takes place before a fracture occurs.

Using these concepts, traumatic spinal instability can be categorized according to the underlying pathophysiological mechanisms. When an axial force vector is applied anterior to the IAR, a compression fracture occurs due to isolated failure of the anterior column.

When the axial force vector is precisely directed over the IAR, no rotation occurs. In this situation, if the stress exceeds the ultimate tensile point of the vertebral bone, failure of both middle and anterior columns occurs, resulting in a burst fracture.

If the axial force vector is directed posterior to the IAR (hyperextension), fractures of laminae and facet joint may result. This is more common in the cervical spine due to its lordotic curvature.

Pure distraction forces are rarely applied to the spine. Distraction-flexion force vectors are composite vectors with components in the superior and anterior orientation in the sagittal plane, generally associated with seat-belt deceleration injuries of the thoracolumbar spine. The vertical (distractive) component of the vector is applied posterior to the IAR, while the flexion component is directed superior to the IAR, resulting in rupture of the posterior ligamentous complex and the middle column. The anterior column remains intact, acting as a hinge. In this type of injury, if the orientation of the vector is such that the flexion component is stronger and is directly applied to the IAR, a true Chance fracture may occur, consisting of a horizontal shearing fracture across the pedicles and/or vertebral endplates.

With even larger flexion force vectors, a fracture-dislocation may occur with failure of all 3 column and bilateral jumped or fractured facets. If a rotational vector (twisting moment) is also present in the axial plane and the flexion vector is not overwhelming, a unilateral jumped facet may result.

A. Compression fracture; B. Burst fracture; C. Hyp A. Compression fracture; B. Burst fracture; C. Hyperextension injury to lamina and facets; D. Flexion-distraction (seatbelt) ligamentous injury and Chance fracture; E. Shear fracture-dislocations.

Although these biomechanical concepts are often discussed in the context of traumatic instability, they can be extended to other forms of instability as well. Furthermore, these principles are commonly applied when devising fusion and instrumentation constructs to treat specific instances of spinal instability. For instance, interbody bone grafts and cages are best applied as distraction constructs applied in the region of IAR. Pedicle screw constructs can act as cantilever beams, shifting the IAR to the rod-screw interface. Consideration of IAR is of crucial importance in 3-point-bending constructs (eg, universal hook, wire, screw, rod systems used for thoracolumbar posterior instrumentation), where application of compressive and distractive forces can have significant effects on spine contour.

Example of application of biomechanical principles Example of application of biomechanical principles to spine surgery. Insertion of special pedicle screws (Schanz screws) pivoting on rod transfers instantaneous axis of rotation (IAR) to screw-rod interface. Compression of proximal end of screws produces distraction-reduction of vertebral burst fracture. If posterior longitudinal ligament is intact, retropulsion is corrected by ligament taxis. Image courtesy of Synthes, Inc.

Biology of fusion

For fusion to succeed, osteoprogenitor cells must differentiate into osteoblasts, populate the fusion matrix, survive in the fusion environment, and deposit bone. Many local and systemic host factors and graft properties affect these processes. Graft material may have osteoconductive, osteoinductive, or osteogenic properties.

Osteoconduction refers to the capacity of the graft to serve as a matrix or scaffolding for infiltration of bone cells and supporting neovascular network. Allogeneic, autologous, and synthetic bone matrices made of hydroxyapatite or coral are osteoconductive.

Osteoinduction refers to the capacity of bone to direct differentiation, migration and attachment of osteoprogenitor cells. Many positive and negative osteoinductive influences exist. Bone morphogenic protein, a member of the transforming growth factor-β (TGF-β) family, induces differentiation of mesenchymal cells into osteoblasts. [10] It is found naturally in the bone fusion environment and is available in recombinant form for clinical use. Compressive forces applied to the bone graft also promote increased bone deposition, accounting for the greater success of interbody bone grafts versus onlay bone grafts. Application of a direct electrical current to bone also has an osteoinductive influence, [11] a phenomenon that is put to use by implantation of a bone stimulator in cases at high risk for pseudarthrosis.

Osteogenic property refers to the capacity of bone graft to initiate fusion by providing live osteoprogenitor cells. Only autologous bone graft has this property.

In addition to osteogenesis, autologous bone graft provides osteoinduction and osteoconduction, thus providing the ideal graft material. A corticocancellous autograft, such as a tricortical iliac crest autograft, is capable of providing structural support as an interbody implant in addition to the above-mentioned favorable properties. The only drawback of using autograft material is the potential for donor site complications associated with graft harvest.

The host factors that adversely affect fusion include malnutrition, corticosteroid use, irradiation, neoplastic disease, diabetes, local infection, osteoporosis, and smoking. Of these, smoking is the most prevalent correctable risk factor. There is abundant experimental and clinical [12] evidence documenting the adverse effects of smoking on bone healing and fusion.

Finally, immobilization of the target motion segment has been shown to significantly enhance the success of fusion. [13] This is best accomplished by instrumentation. In absence of instrumentation, fusion should be supported by external bracing until it solidifies.



The clinical manifestations of spinal stability fall into 3 categories, as stated in the definition of instability:

  • Neurological deficit due to cord, cauda equina, or nerve root compression
  • Pain
  • Incapacitating deformity


Indications for fusion in acute overt instability

Conditions that result in acute overt instability require stabilization, either internally (by fusion) or externally (by reduction and bracing). In traumatic injuries, if instability is due to a fracture rather than ligamentous rupture, if the fracture fragments are (or can be reduced to be) in contact and in near-anatomical alignment, and if there is no significant neural compression, an external brace (eg, halo, collar, thoracic lumbar sacral orthosis [TLSO] brace) is tried until the fracture heals. In all other circumstances and in cases where bracing has failed, fusion is indicated. Tables 1 and 2 summarize treatment algorithms and indication for fusion in cervical, thoracic, and lumbar spine trauma.

Table 1. Traumatic Instability of Cervical Spine (Open Table in a new window)



Type/Issue Treatment
C1 Jefferson fracture

(axial loading)

1. Isolated -->

2. With transverse ligament rupture -->

3. Widely diastatic -->

4. With odontoid fracture -->

1. Hard collar

2. Halo

3. Consider occiput-C2 fusion

4. Treat according to odontoid fx

C1-2 Rotatory subluxation

(twisting moment)

1. Children, URI -->

2. Adults, tumor, trauma, infection -->

1. Bedrest, analgesics, halter traction, soft collar

2. Traction, hard collar, halo, or C1-2 fusion depending on cause and duration

Odontoid fracture

(flexion in young, extension in old)

1. Type 1 -->

2. Type 2, < 6 mm displaced -->

3. Type 2, >6 mm displaced or chronic or type 2A -->

4. Type 3 -->

1. If no atlanto-occipital instability, collar x 3 mo

2. Halo x 3-6 mo

3. C1-2 fusion or odontoid screw

4. Halo x 6 mo

C2 Hangman fracture


1. Pars approximated->

2. Pars separated, reducible -->

3. Pars separated, not reducible -->

1. Hard collar x 3 mo

2. Reduce in extension, then halo x 3 mo

3. C2-3 fusion

Unilateral jumped facet

(flexion + rotation)

1. Reducible -->

2. Not reducible -->

3. With facet fracture -->

4. With disc herniation-->

1. Reduce and halo x 3 mo

2. Open reduction and posterior fusion

3. Open reduction and posterior fusion

4. Anterior decompression, open reduction, and anterior fusion

Bilateral jumped facet


1. Reducible, without disc herniation -->

2. Not Reducible, without disc herniation-->

3. With disc herniation-->

1. Closed reduction, then posterior fusion

2. Open anterior or posterior reduction and fusion

3. Anterior discectomy, reduction and fusion

Subaxial spine axial loading injuries

(axial +/- flexion)

1. Simple compression fracture -->

2. Burst fracture +/- tear drop fx -->

3. Burst + posterior column fracture -->

1. Hard collar

2. Anterior corpectomy and fusion

3. Anterior corpectomy and fusion

(+/- posterior fusion)

Clay shoveler fracture


Always stable Soft collar and analgesics
Anterior avulsion fracture


Always stable Soft collar and analgesics

Table 2. Traumatic Instability of Thoracic and Lumbar Spine (Open Table in a new window)

Fracture Denis Columns Involved Treatment
Compression fracture Anterior column Bracing (note that >50% vertebral body height loss or Cobb angle >30 degrees predict worsening of kyphosis)
Compression fracture with splaying of spinous processes Anterior and posterior columns Posterior instrumented fusion
Stable burst fracture

(preserved posterior longitudinal ligament)

Anterior column and part of middle column If no neural compromise, treat with TLSO brace

If canal stenosis present, retropulsed fragment may be reduced by ligamentous taxis in distraction

with posterior instrumented fusion

Unstable burst fracture Anterior and middle columns with significant retropulsion,

or all 3 columns

Anterior decompression and instrumented fusion
Flexion-distraction seat belt injury (ligamentous) Middle and posterior columns Posterior reduction and instrumented fusion
Chance fracture (osseous) 2 or 3 columns but with good bone contact TLSO brace
Shear fracture dislocation 3 columns Instrumented fusion, anterior, posterior, or both

When overt instability is produced by a tumor, indications for surgery depend on the patient's life expectancy, physical condition, extent of cord compression, responsiveness to radiation and chemotherapy, number of motion segments involved by tumor, and severity of pain. The ideal candidate for decompression and fusion is a patient with limited systemic and spinal neoplastic disease who presents with an acute pathological fracture with incomplete cord compromise.

Infections of the spine, if discovered early, may produce no neural compromise or instability and may be treated by antibiotics alone. However, advanced infections of the discs and vertebral bodies are highly destructive and destabilizing, requiring debridement/decompression and fusion, either simultaneously or in separate sessions.

Indications for fusion in chronic overt instability

Chronic overt instability is initially managed conservatively (analgesics, anti-inflammatory drugs, physical therapy, bracing). If and when the patient fails to respond to conservative management or if significant neurological compromise exists, fusion is indicated.

Indications for fusion in anticipated instability

Surgical removal of two columns of the spine (or removal of one column when another is known to be deficient), radical removal of one facet joint (see image below), or partial substantial removal of both facet joints in one motion segment would be expected to produce instability. In these cases, it is prudent to consider fusion at the time of the original surgery.

Example of anticipated instability. Figure A shows Example of anticipated instability. Figure A shows large mass affecting right C3-4 facet joint and lateral masses in patient with severe right-side neck and shoulder pain; Figures B and C show complete resection of tumor and simultaneous C3-4 anterior fusion to circumvent anticipated iatrogenic stability produced by radical resection of facet and lateral masses.

Indications for fusion in covert instability

Like chronic overt instability, covert instability is initially managed conservatively, but with a much higher threshold for abandoning conservative treatment in favor of fusion. Isolated spondylolysis without spondylolisthesis and spondylolisthesis without dynamic instability are typically treated conservatively with physical therapy and epidural steroid injections for at least 3-6 months. If back pain exists without radicular symptoms, greater effort is made to avoid surgery. The patient must quit smoking and demonstrate the ability to limit the intake of narcotics. With appropriate patient selection, good results can be achieved with fusion when conservative treatment has failed. In symptomatic spinal stenosis without spondylolisthesis, decompression alone is the treatment of choice, but in spinal stenosis with degenerative spondylolisthesis of significance, fusion improves outcome. [14, 15]

Much more controversial is the treatment of that subcategory of covert instability that is known as microinstability or dysfunctional motion segment. Here, an abnormal disc or facet joint is presumed to be the pain generator. Provocative discography and facet injections are often used in this setting to "locate" the pain generator. The idea is that fusion, by eliminating motion across the dysfunctional motion segment, may alleviate the pain. This controversy and the relevant recommendations of the American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Section on Disorders of Spine and Peripheral Nerves are explored in greater detail in Outcome and Prognosis section.


Relevant Anatomy

Regional variations in vertebral anatomy affect the incidence and consequences of spinal instability in different parts of the spine and dictate the surgical means by which the spine can be stabilized.

Vertebral body size increases as one descends the spine, accompanied by a corresponding increase in axial load-bearing capacity of the vertebrae. The greater cancellous-to-cortical bone ratio in the vertebral body compared to the posterior vertebral elements makes it more susceptible to neoplastic and infectious diseases, while its relationship to the IAR makes it more susceptible to compressive injuries. The relative preponderance of these disorders anterior to the spinal cord makes their surgical management more challenging, often requiring an anterior surgical approach to the spine. On the other hand, the large surface area and volume of the vertebral body make it an excellent target for insertion of screw/plate systems, which can be used to stabilize every segment of the subaxial spine.

Facet joints have a transverse orientation in the cervical spine and gradually acquire a more sagittal orientation throughout the thoracic and upper lumbar spine. They then become more coronally oriented as one descends the lumbar spine. The transverse orientation of the facet joints and the loose facet capsules in the cervical spine provide for relatively free movements of the neck in all three planes and do not protect the cervical spine against flexion injuries. In the thoracolumbar junction, the sagittal orientation of the facet joints and the strong capsular ligaments provide for greater movement in the sagittal plane than in other directions. This facet orientation and the transitional location of the thoracolumbar spine between the rib cage-stabilized thoracic spine and the more robust lumbar spine make the thoracolumbar junction more susceptible to flexion injuries.

The more coronal orientation of the L5-S1 facet joints compared to the L4-5 facets accounts for the lower incidence of degenerative spondylolisthesis at L5-S1, in spite of the biomechanically disadvantaged angle of the lumbosacral junction. In contrast, isthmic spondylolisthesis, where the presence of spondylolysis bypasses the resistance of facet joints against translation, is more frequent at L5-S1.

The spinal canal is narrowest in the thoracic spine. On the other hand, the thoracic spine is stabilized by the ribcage, making it relatively immune to degenerative instability and increasing its resistance to traumatic instability. Consequently, if the force vector is great enough to overcome the stability of thoracic spine and produce a fracture/dislocation, the likelihood and severity of spinal cord injury would be greater than elsewhere in the spine.

The pedicles in the cervical spine are quite narrow, short, acutely oriented, and juxtaposed to the transverse foramina (of the vertebral artery), making them relatively undesirable for screw insertion. In contrast, the large size, strength, and favorable cylindrical anatomy of the pedicles in the lumbar spine makes them ideal for screw insertion. The pedicle screws at different segments are then linked by rods to stabilize the spine. The pedicles acquire a relatively sagittal orientation in the thoracic and upper lumbar spine, but then point inward again as one approaches the sacrum, a fact that has to be taken into account when inserting pedicle screws. In the thoracic spine, the pedicles have a narrow transverse diameter, a slight downward angle, and are located next to the narrow thoracic spinal canal.

Because of these anatomical considerations, wires and hooks have been used more than screws to anchor rods against the thoracic spine, necessitating long instrumentation constructs to stabilize a short segment of instability ("rod long, fuse short"). Increasingly, screws are used in the thoracic spine to create shorter and stronger instrumentation constructs. In this setting, it is imperative that screws of appropriate diameter be selected based on preoperative CT studies and that breach of the medial pedicle wall be avoided, erring toward the laterally located and protective costovertebral articulation, if necessary. On the other hand, the relatively generous sagittal diameter of thoracic pedicles and the smaller size and lesser functional importance of thoracic nerve roots make screw misdirection in the sagittal plane less costly in the thoracic spine than in the lumbar spine.

Cervical vertebrae have anatomical structures not found elsewhere in the spine – the lateral masses. Juxtaposed between the pedicles and the lamina and delimited by the articular surfaces of the adjacent facet joints, the paired lateral masses are satisfactory targets for screw insertion. Lateral mass screws at adjacent segments are linked by plates or rods to stabilize the cervical spine.

Laminae, spinous processes, and transverse processes can be used as anchor points for wires and hooks connected to rods to form three-point-bending instrumentation constructs. Alternatively, these structures can be wired to each other at different segments to produce tension band constructs. In general, these types of constructs provide less stiffness than screw/rod or screw/plate systems.

Comparison of vertebral anatomy in cervical, thora Comparison of vertebral anatomy in cervical, thoracic, and lumbar spine. Note variation in anatomy and size of pedicles.


Absolute contraindications to fusion are relatively uncommon and include the following:

  • Diffuse multilevel neoplastic disease such that no adjacent normal segments exist for engagement of instrumentation
  • Severe osteoporosis such that the bones would not support instrumentation and fusion would not be expected to solidify in absence of instrumentation
  • Infection of soft tissues adjacent to spine or epidural infection that has not spread to the vertebral bones or discs, in which case the fusion construct would be at risk for infection (see below for established discitis/osteomyelitis)

Relative contraindications to spinal fusion include the following:

  • Smoking
  • Malnutrition
  • Systemic infection
  • Chronic hypoxemia
  • Severe cardiopulmonary disease
  • Severe depression, psychosocial issues, and secondary gain issues

As always, the contraindications to surgery have to be weighed against the risks of not performing the operation in each particular situation. For instance, smoking and severe depression may be contraindications to fusion in a patient with back pain and disc degeneration, but should not deter the surgeon from fusing an unstable cervical spine fracture.

Importantly, an active spine infection (discitis/osteomyelitis) does not necessarily constitute a contraindication to fusion and instrumentation. To the contrary, advanced spine infections exert severe destabilizing effects on the spine, often requiring stabilization at the time of debridement and decompression. In this setting, careful clinical, laboratory, and radiographic follow-up are essential as the patient receives prolonged intravenous antibiotic treatment (for at least 6 wk) to confirm eradication of the infection. Worsening pain or neurological deficit, persistent fever, leukocytosis, or bacteremia and persistently elevated erythrocyte sedimentation rate signal the possibility of persistent infection.

Similarly, radiographic evidence of loosening of screws or CT scan/MRI evidence of increased bone destruction should be further investigated. However, persistent and stable vertebral enhancement on MRI does not necessarily indicate persistent infection, as this finding can lag behind microbiological cure. Radionucleotide bone scan lacks specificity in this setting, but a tagged-WBC scan may be more useful. When in doubt, a CT-guided biopsy/aspiration of the region can help confirm the possibility of persistent infection, which would then be treated with reoperation.

Loosening of this infected pedicle screw is eviden Loosening of this infected pedicle screw is evidenced by radiolucent halo (arrows) surrounding screw.
In this patient with T7-8 discitis, vertebral enha In this patient with T7-8 discitis, vertebral enhancement on MRI persisted 8 weeks after clinical and microbiologic cure.