Thoracolumbar Fractures




Classification



Andrei F. Joaquim
Alpesh A. Patel

Introduction


The thoracolumbar spine represents the most common site of fractures in the spine. It is comprised of injuries to the thoracic spine (T1-T10), thoracolumbar junction (T11-L2), and lumbar spine (L3-L5).


Classification of spinal injuries is a critical step. It provides a common language for identification and comparison of injuries. This common language is fundamentally important in clinical care, education, and research as it allows for valid comparisons between physicians, institutions, and nationalities in the treatment of thoracolumbar spine trauma (TLST).


An ideal classification system must be precise, accurate, and valid. Precision or reliability refers to the ability of repeat application of the classification to the same condition with similar results. It is commonly evaluated with the percent of agreement. Accuracy is the truth to which, compared to a reference standard, the classification defines the image. Sensitivity, specificity, and positive and negative predictive values are measures of the accuracy. Besides precision and accuracy, a measurement must be valid in the clinical context to be useful. A classification must also be predictive of the patient’s outcome and aid in the comparison of potential forms of treatment.


Historical Review


In 1934, Böhler grouped spine fractures into five injuries according to their morphology, based on plain radiographs. In 1949, Nicoll proposed a classification of thoracic and lumbar fractures into four types: (1) anterior edge fracture, (2) lateral edge fracture, (3) fracture dislocation, and (4) isolated fractures of the neural arch. The author postulated that fracture stability was associated with the integrity of the interspinous ligament. Fractures, therefore, could be labeled as stable or unstable based on their integrity. This concept was critically important as it was the first description of the role of the posterior ligamentous complex (PLC) in spinal stability.


In 1970, Holdsworth introduced a new classification scheme based on the concept of columns. He proposed that there were two columns responsible for vertebral stability: an anterior column, composed by the vertebral body, intervertebral disc, and anterior and posterior longitudinal ligaments; and a posterior column, composed by the facet joints, interspinous, supraspinous, and yellow ligaments. He also stated that spinal stability required the integrity of the posterior column. Four basic mechanisms of injuries were described: compression, flexion, extension, and flexion-rotation. Holdsworth also described the term “burst” as a fracture secondary to axial overload with consequent herniation of the nucleus pulposus into the superior vertebral endplate. He considered this injury stable.


In 1977, Louis modified Holdsworth’s concept of two columns to three columns: one anterior and two posterior. The anterior column was formed by the vertebral body and disc, whereas the two posterior columns were formed by the facet joints and ligaments of each side. He also proposed that the lamina and the pedicles were responsible for additional stabilization of the vertebrae. Another important point in the Louis system was the graduation of the stability based on a nominal system; he postulated that there were degrees of instability, within a continuum between a stable and an unstable injury. Based on morphologic and mechanical criteria, values were attributed to injuries: 2 points for loss of substance in the columns; 1 point for vertical spine injuries; 0.5 point for incomplete injuries of the pedicles, vertebral body, or lamina; 0.25 point for fractures in the transverse or spinous process. Injuries with 2 or more points were considered unstable. For the first time, treatment was proposed based on a numerical score.


In 1983, based on evaluation of 100 cases of thoracolumbar fractures using then newly available computed tomography (CT), McAfee and colleagues proposed six different injuries. They also proposed that burst fractures can be stable when anterior and middle columns fail because of a compressive load with no loss of integrity of the posterior elements and unstable when the posterior elements are disrupted. This disruption can lead to failure in compression, lateral flexion, or rotation, but most likely results in posttraumatic kyphosis and the development of neurologic symptoms.


In 1983, Denis modified the column models proposed by Holdsworth and Louis. The spine was again divided into three columns: (1) an anterior column (anterior half of the vertebral body and intervertebral disc, within the anterior longitudinal ligament); (2) a middle column (posterior half of the vertebral body and intervertebral disc plus the posterior longitudinal ligament); and (3) a posterior column (supraspinous, interspinous, and yellow ligament plus the facet joints) ( Fig. 35A-1 ). In his theory, Denis proposed that an isolated posterior injury would not compromise stability. Injury to two columns would be necessary to make the spine “unstable.” Denis also classified spinal injuries in two groups: major and minor. This latter minor group includes transverse and spinous process fractures, as well as pars interarticularis fractures. Major injuries were then classified in five types: A, B, C, D, and E. The burst fracture concept was redefined as an injury to the anterior and middle columns and classified as unstable, as two columns were injured. Although widely used, the Denis system was criticized for not being comprehensive, for having low reliability and reproducibility, and for the theoretical nature of the middle column, in that it is not an anatomic entity and, therefore, characterization of middle column injuries remained difficult.




Figure 35A-1


Denis’ three-column model of the spine. The middle column is made up of the posterior longitudinal ligament, the posterior annulus fibrosus, and the posterior aspects of the vertebral body and disc.


In 1990, White and Panjabi defined spinal injuries based on the concept of stability, which is the ability of supporting elements of the spine to resist physiologic loads so as to prevent neurologic injury, deformity, or pain. This descrip­tor provides continuing clarity of the clinical definition of spinal stability and remains a critical determinant of treatment.


McCormack and colleagues proposed a point system in 1994 known as load-sharing classification (LSC) to distinguish three-column spinal fractures according to the amount of damaged vertebral body, the spread of the fragments, and the amount of corrected kyphosis after short-segment instrumentation. Their system was an attempt to predict screw breakage in short-segment reconstruction and to identify spinal fractures requiring additional anterior reconstruction.


The classification system developed by Magerl and coworkers, also known as the Arbeitsgemeinschaft für Osteosyn­thesefragen (AO) Classification of Fractures ( Table 35A-1 ), is probably the most widely discussed system in 1990s and early 2000s. It was a pathomechanistic classification based on a review of 1445 plain radiograph fractures. The system considered prognostic aspects regarding healing and patient recovery. Injuries are primarily classified according to the main mechanism of injury as type A (axial force injuries, including compression and burst fractures), type B (distraction), and type C (rotational).







    • Type A: Injuries secondary to compression fractures, in the vertebral body, such as compression and burst fractures, without injury to the PLC. Small bone injury in the posterior elements can be present.



    • Type B: Injuries resulting in transverse disruption, with anterior or posterior distraction.



    • Type C: The most severe injuries, with anterior and posterior element disruption and rotation.




TABLE 35A-1

MAGERL CLASSIFICATION OF SPINE INJURIES (AO CLASSIFICATION SYSTEM)
















Type Characteristic
Type A Compression and burst fractures (no PLC injury)
Type B Anterior and/or posterior element injuries with distraction (vertical forces)
Type C Anterior and posterior element injuries with rotation

PLC, Posterior ligamentous complex.

Source: Magerl F, Aebi M, Gertzbein SD, et al: A comprehensive classification of thoracic and lumbar injuries, Eur Spine J 3:184–201, 1994.


These injuries are then subclassified into detailed and numerical groups and subgroups, with more than 50 subtypes. The severity of the cases increases from A to C, also increasing incrementally according to groups and subgroups.


The Magerl system is frequently criticized for its high degree of complexity, decreasing its reliability, and also for not considering the neurologic status in the decision-making process. Although beneficial for research, these factors limit its application in clinical practice and education. It is also important to mention that the system was proposed based on plain radiographs, without considering the potential benefits of modern high-resolution CT scanning with reconstruction and magnetic resonance imaging (MRI) in defining spinal injuries. Even considering its potential pitfalls, the AO system is still used worldwide.


Thoracolumbar Injury Classification System


Considering this entire historical context, in 2005 the Spine Trauma Study Group proposed a new system to help surgeons classify and treat thoracic and lumbar spine fractures. The system is based on three major descriptors, all of them associated with the treatment and prognosis of these injuries: (1) injury morphology, (2) integrity of the PLC, and (3) neurologic status.


The system is known as Thoracolumbar Injury Classification System and Severity Score (TLICS) and was developed to help surgeons guide treatment ( Table 35A-2 ). According to the obtained score, conservative or surgical treatment can be proposed. Its main advantage is its quantification of the neurologic status and also its assessment of the integrity of the PLC and the effect of these critical factors on the decision-making process.



TABLE 35A-2

THORACOLUMBAR INJURY CLASSIFICATION SYSTEM AND SEVERITY SCORE (TLICS)























































Variable Points
Injury Morphology
Compression 1
Burst +1
Translation/rotation 3
Distraction 4
Neurologic Status
Intact 0
Nerve injury 2
Cord, conus medullaris
Incomplete 3
Complete 2
Cauda equina 3
PLC Integrity
Intact 0
Indeterminate * 2
Injured 3

PLC, Posterior ligamentous complex.

* Indeterminate status is attributed to patients without evident disruption on computed tomography (CT) scan reconstructions (without clear dislocation) but with suggested ligamentous injury at the short tau inversion recovery (STIR) or T2-weighted magnetic resonance imaging (MRI) sequence.



After all individual characteristics are classified and scored, the summation of each variable leads to a final score. When three or fewer points are obtained, conservative treatment is proposed. Five or more points suggest that surgical treatment is probably the best treatment option. Patients with four points can be treated conservatively or surgically, according to surgeon’s preference. Other variables can also influence the final treatment option, such as body habitus, obesity, comorbidities, patient age, systemic injuries, patient preference, institutional or health system capacity, among many others ( Figs. 35A-2 and 35A-3 ).




Figure 35A-2


A 22-year-old man presented after a car accident and an L3 translational injury with no neurologic status. A and B, Axial computed tomography (CT) scan showing vertebral body and laminar fracture. A sagittal ( C and D ) and coronal ( E ) CT scan reconstruction showing a rotational injury between L2 and L3, with spinal translation in both planes. Diastasis of facet joints can be seen ( D ). This injury can be classified as type C, according to the Arbeitsgemeinschaft für Osteosynthesefragen (AO) classification system. Applying the Thoracolumbar Injury Classification System and Severity Score (TLICS), we have 3 points for morphology plus 3 points for posterior ligamentous complex (PLC) plus 0 for neurologic status, giving a total of 6 points. Surgical treatment was proposed, with L1 to L5 posterior instrumentation, as shown in F and G.



Figure 35A-3


A 57-year-old man sustained a fall from a height with an L1 compression and an L3 burst fracture. A and B, There is no facet joint diastasis or any rotation or distraction evident in sagittal computed tomography (CT) reconstruction. C, An axial CT scan of L3 showing injury of the posterior body wall and canal compression. D, A sagittal T2-weighted magnetic resonance imaging (MRI). There is no posterior ligamentous complex (PLC) injury based on CT and MRI. The patient was neurologically intact. L1 and L3 are Arbeitsgemeinschaft für Osteosynthesefragen (AO) type A injuries (compression and burst fractures). The Thoracolumbar Injury Classification System and Severity Score (TLICS) for L1 fracture is 1 point for morphology plus 0 for PLC plus 0 for neurologic status, giving a total of 1 point. The TLICS for L3 was 2 points for morphology plus 0 points for PLC plus 0 for neurologic status, giving a total of 2 points. The patient was successfully treated nonsurgically.


TLICS has been well accepted by the scientific community. Some studies suggested good reliability and reproducibility, and also attest that it is an important tool for medical education.


Updated AO Classification


Vaccaro and colleagues as the AOSpine Spinal Cord Injury and Trauma Knowledge Forum recently published an updated thoracolumbar (TL) spine injury classification system. The new AOSpine TL injury classification system, similar to the TLICS system, defines three basic parameters: morphologic classification of the fracture, neurologic status, and clinical modifiers. Fracture morphology is defined as type A (compression injuries), type B (tension band injuries), and type C (translation/displacement). Neurologic status is defined as N0 (intact), N1 (transient deficit, resolved), N2 (radiculopathy), N3 (incomplete spinal cord or cauda equina injury), and N4 (complete spinal cord injury, American Spinal Injury Association [ASIA] grade A). Finally, two clinical modifiers were reported: M1, fractures with indeterminate injury to tension band; and M2, patient-specific comorbidities. Fracture morphology is further defined by subgroups of A0 to A4 and B1 to B3 ( Table 35A-3 ).



TABLE 35A-3

SUMMARY OF THE NEW AOSPINE THORACOLUMBAR TRAUMA CLASSIFICATION SYSTEM































































Fracture Morphology
A Compression
A0 No injury/process fracture
A1 Wedge/impaction
A2 Split/pincer
A3 Incomplete burst
A4 Complete burst
B Tension Band Injuries
B1 Posterior transosseous disruption
B2 Posterior ligamentous disruption
B3 Anterior ligamentous disruption
C Translation/Displacement
Neurologic Status
N0 Intact
N1 Transient, resolved
N2 Radiculopathy
N3 Incomplete spinal cord or cauda equina injury
N4 Complete spinal cord injury (ASIA grade A)
Clinical Modifiers
M1 Indeterminate tension band
M2 Patient-specific comorbidities (AS, DISH, osteopenia, etc.)

AS, Ankylosing spondylitis; ASIA, American Spinal Injury Association; DISH, diffuse idiopathic skeletal hyperostosis.


The authors reported overall moderate-to-fair interobserver reliability and substantial-to-excellent intraobserver reliability. With the addition of the morphology subtypes, however, reliability decreased. The new AO system has not been tested outside of the describing authors and, therefore, its generalizability to other surgeons, residents, and students remains unknown. Furthermore, the new system has not been clinically validated in patient care. Future studies may address these informational gaps, providing a more complete assessment of the new system.


Reliability of Classification Systems


Reliability refers to the extent to which repeated measurements of the same situation agree with each other. Potential sources of variation include the patient, the physician, and the radiologic instrument used (such as CT scan vs. MRI). The variability of the physician can be divided into intrarater (when the reliability is accessed by the same physician) or interrater (when accessed by a different health care provider). As a practical point of view, interrater reliability is the most important evaluation to consider a classification system reproducible and potentially useful.


Reliability is generally measured by Cohen’s kappa coefficient, a statistical instrument to quantify agreement. Generally, the magnitude of the agreement is classified according to the kappa value ( Table 35A-4 ). Table 35A-5 presents the reliability of the most useful classification systems.



TABLE 35A-4

MAGNITUDE OF AGREEMENT ACCORDING TO COHEN’S KAPPA COEFFICIENT

























Kappa Value Magnitude of the Agreement
<0 Chance
0.01–0.20 Slight
0.21–0.40 Fair
0.41–0.60 Moderate
0.61–0.80 Substantial
0.81–0.99 Almost perfect


TABLE 35A-5

INTERRATER RELIABILITY OF THE MOST COMMON CLASSIFICATION SYSTEMS ACCORDING TO COHEN’S KAPPA COEFFICIENT IN DIFFERENT STUDIES


















































System Interrater Reliability Range (Kappa Score)
Dennis System For major fracture subtype 0.52–0.60
For entire classification system 0.39–0.45
Magerl System For the three main types 0.33–0.42 (studies using plain radiographs, CT, or MRI)
TLICS Morphology 0.608
PLC status 0.641
Neurologic status 0.91
Total score 0.576–0.74
Updated AO
Type A 0.72
Type B 0.58
Type C 0.70

AO, Arbeitsgemeinschaft für Osteosynthesefragen; CT, computed tomography; MRI, magnetic resonance imaging; PLC, posterior ligamentous complex; TLICS, Thoracolumbar Injury Classification System and Severity Score.


Morphology and Classification System


Morphology is one of the most crucial characteristics in a classification system for traumatic spine injuries. Some basic morphologic characteristics can be easily described, such as fracture line description and vertebral displacement. Some fracture morphologies are related to spinal instability and can guide treatment. Morphology characteristics are generally used indirectly to describe the severity of a spinal injury. The main morphology described in thoracolumbar trauma are briefly presented in the following sections.


Compression Fractures


Compression fractures are generally considered stable injuries; they exhibit minimal body height loss (<10%) and kyphosis (<25 degrees) ( Fig. 35A-4 ). By definition, compression fractures do not involve the posterior wall of the vertebral body. Compression fractures most commonly involve the superior and anterior endplate resulting in vertebral body wedging. Other patterns are the anterior body only, the pincer fracture where a fracture line divides the anterior and posterior vertebral body, and less commonly, an inferior endplate fracture. An important component of compression fractures is the maintenance of the posterior osseous ligamentous complex.




Figure 35A-4


Comprehensive classification of type A spinal injuries. The three categories of type A fractures include impaction injuries (A1), of which wedge fractures are most commonly seen; split fractures (A2), of which the pincer fracture is the typical injury; and the burst fracture (A3).

(Redrawn by permission from Gertzbein SD: Classification of thoracic and lumbar fractures. In Gertzbein SD, editor: Fractures of the thoracic and lumbar spine, Baltimore, 1992, Williams & Wilkins.)


Burst Fractures


Burst fractures are characterized by fractures and loss of height of the anterior and posterior portions of the vertebral body. The hallmark of the burst fracture is the retropulsion of the posterior wall and canal compromise. This may lead to neurologic deficits. Centrifugal extrusions of the bone fragments are generally presented. The pedicles are expanded apart, which is best seen as widening of anteroposterior or coronal plane reconstructions. Like compression fractures, several patterns of body and posterior involvement have been described that might affect treatment decisions. Most commonly, the superior endplate is fractured, whereas a more unstable injury is when the fracture comminution extends to involve both superior and inferior endplates.


Magerl and colleagues proposed that burst fractures have an intact PLC. However, the authors of TLICS proposed that some burst fractures can have PLC injury, having an increasing risk of instability. Burst fractures can be associated with laminar and spinous process fractures.


Perhaps the greatest impact the TLICS system can have is in the evaluation and treatment of burst fractures. The TLICS system does not account for some classically described characteristics of burst fractures. Factors such as loss of vertebral body height, segmental kyphosis, and canal compromise, despite their widespread use, have no evidence to support their importance in managing patients without neurologic deficits. Studies have examined the impact of these classic radiographic measures on patient outcome and have shown no relationships with only neurologic function consistently predicting patient outcomes. As an example, kyphosis greater than 30 degrees and/or 50% of vertebral body height loss were indirect radiologic measures of PLC disruption, such as substantial canal compromise. Based on these historical analyses, Radcliff and coworkers studied the relation of loss of vertebral body height, kyphosis, and canal compromise observed on CT scan with the PLC status according to MRI findings. They reported that these factors did not correlate with PLC injury and the MRI should be used when there is clinical concern. However, there is lack of consistent evidence on the utility of MRI in the evaluation and management of spinal fractures. As such, despite the proven utility of the TLICS system, the controversy regarding the treatment of burst fractures remains, especially in the evaluation of the PLC status. Over time, with additional evidence and clinical utilization of the TLICS system, consistency in the treatment of stable burst fractures can be achieved with the betterment of patient treatment and outcomes.


Flexion-Distraction Injuries


Flexion-distraction injuries are a flexion injury of the spine, characterized by a compression injury to the anterior portion of the vertebral body and a transverse fracture through the posterior elements of the vertebra and the posterior portion of the vertebral body ( Fig. 35A-5 ).




Figure 35A-5


Flexion-distraction injury at T11-T12 demonstrates small anterior compression injury but wide posterior splaying consistent with posterior distraction.


The mechanism of injury is rotation about a fulcrum located within the vertebral body, so that the middle and posterior columns fail in tension and the anterior column fails in compression. Its association with extreme forward flexion in automotive accidents results in the literature name of “seatbelt” fractures. The Chance fracture variant occurs when all three columns fail under tension often involving only bony structures ( Fig. 35A-6 ). Various patterns of flexion-distraction and Chance fractures have been described and are differentiated by combinations of bony, soft tissue, and facet articulation involvement.




Figure 35A-6


Denis’ classification of flexion-distraction injuries. These may occur at one level through bone ( A ), at one level through the ligaments and disc ( B ), at two levels, with the middle column injured through bone ( C ), or at two levels with the middle column injured through ligament and disc ( D ).


Many abdominal injuries especially bowel perforation were described in association with flexion-distraction injuries and lead to a high level of suspicion in patients with this injury morphology.


Hyperextension Injuries


Hyperextension injuries can present with anterior column disruption through the disc or the vertebral body, secondary to severe spine hyperextension. These injuries are frequently described in patients with diffuse idiopathic skeletal hyperostosis (DISH) and ankylosing spondylitis (AS) and are considered unstable. They can have or not have PLC involvement, increasing the degree of instability. Sometimes it is possible to see a fracture line crossing the syndesmophytes of the ankylosed spine.


Fracture-Dislocation


Fracture-dislocations are the most severe spinal cord injuries and are commonly associated with neurologic deficits. These injuries have vertebral body displacement, complex fractures, and joint subluxation and rotation, in a multitude of patterns with displacement beyond the normal range ( Fig. 35A-7 ). Compression and burst fractures can be found in association with this complex group of injuries.




Figure 35A-7


Denis’ classification of fracture-dislocation of the spine. A, Type A is a flexion-rotation injury, occurring either through bone or through the disc. There is a complete disruption of all three columns of the spine, usually with the anterior longitudinal ligament remaining the only intact structure. Commonly, this is stripped off the anterior portion of the vertebral body below. These injuries are usually associated with fractures of the superior facet of the more caudal vertebra. B, Type B is a shear injury. The type that produces anterior spondylolisthesis of the more cephalad vertebra usually fractures a facet, but the type that causes a posterior lithiasis of the more cephalad vertebra normally does not cause a fracture of the facet joint. C, Type C is a bilateral facet dislocation. This is a flexion-distraction injury but with disruption of the anterior column in addition to the posterior and middle columns. This disruption through the anterior column may occur through either the anterior intervertebral disc or the anterior vertebral body.


Conclusions


The use of classification systems is to provide taxonomy for communication and to guide treatment. In clinical practice, the fracture pattern is assessed and a morphologic name is associated. This is important as each morphologic type may have its own treatment paradigm. Then the severity of injury is determined using systems such as AO and TLICS or for burst fractures, the McCormack system. This is important as injuries vary in the severity even when having the same morphologic type. In the following chapters in this section, the treatment of the various fracture types is discussed.


References


The level of evidence (LOE) is determined according to the criteria provided in the preface.


  • 1. Wang H, Zhang Y, Xiang Q, et. al.: Epidemiology of traumatic spinal fractures: experience from medical university–affiliated hospitals in Chongqing, China, 2001–2010. J Neurosurg Spine 2012; 17: pp. 459-468.
  • 2. Böhler L, editor: Técnica del tratamiento de las fracturas, Barcelona, 1934. Editorial Labor.
  • 3. Nicoll EA: Fractures of the dorso-lumbar spine. J Bone Joint Surg Br 1949; 31: pp. 376-394.
  • 4. Holdsworth FW: Fractures, dislocations and fracture–dislocations of the spine. J Bone Joint Surg Am 1970; 52: pp. 1534-1551.
  • 5. Louis R: Les theories de l’instabilité. Rev Chir Orthop Reparative Appar Mot 1977; 63: pp. 423-425.
  • 6. McAfee PC, Yuan HA, Fredrickson BE, et. al.: The value of computed tomography in thoracolumbar fractures: an analysis of one hundred consecutive cases and a new classification. J Bone Joint Surg Am 1983; 65: pp. 461-473. LOE II
  • 7. Denis F: The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine (Phila Pa 1976) 1983; 8: pp. 817-831. LOE III
  • 8. Aebi M: Classification of thoracolumbar fractures and dislocations. Eur Spine J 2009; 19: pp. S2-S7.
  • 9. Magerl F, Aebi M, Gertzbein SD, et. al.: A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 1994; 3: pp. 184-201. LOE II
  • 10. White AA, Panjabi MM: Clinical biomechanics of the spine. 1978. J.B. Lippincott Philadelphia
  • 11. McCormack T, Karaikovic E, Gaines RW: The load-sharing classification of spine fractures. Spine (Phila Pa 1976) 1994; 19: pp. 1741-1744. LOE II
  • 12. Oner F: Thoracolumbar spine fractures: diagnostic and prognostic parameters [Academic Thesis]. 1999. University of Utrecht Utrecht http://www.library.uu.nl/digiarchief/dip/diss/1885237/inhoud.htm
  • 13. Vaccaro AR, Lehman RA, Hulbert PA, et. al.: A new classification of thoracolumbar injuries: the importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine (Phila Pa 1976) 2005; 30: pp. 2325-2333. LOE II
  • 14. Koh YD, Kim DJ, Koh YW: Reliability and Validity of Thoracolumbar Injury Classification and Severity Score (TLICS). Asian Spine J 2010; 4: pp. 109-117. LOE I
  • 15. Lewkonia P, Paolucci EO, Thomas K: Reliability of the thoracolumbar injury classification and severity score and comparison with the Denis classification for injury to the thoracic and lumbar spine. Spine (Phila Pa 1976) 2012; 37: pp. 2161-2167. LOE I
  • 16. Joaquim AF, Fernandes YB, Cavalcante RC, et. al.: Evaluation of the Thoracolumbar Injury Classification System in thoracic and lumbar spinal trauma. Spine (Phila Pa 1976) 2011; 1: pp. 33-36. LOE I
  • 17. Patel AA, Vaccaro AR, Albert TJ, et. al.: The adoption of a new classification system: time-dependent variation in interobserver reliability of the thoracolumbar injury severity score classification system. Spine (Phila Pa 1976) 2007; 32: pp. E105-E110. LOE III
  • 18. Vaccaro AR, Baron EM, Sanfilippo J, et. al.: Reliability of a novel classification system for thoracolumbar injuries: the Thoracolumbar Injury Severity Score. Spine (Phila Pa 1976) 2006; 31: pp. S62-S69.
  • 19. Vaccaro A, Oner C, Kepler C, et. al.: AOSpine thoracolumbar spine injury classification system: fracture description, neurological status, and key modifiers. Spine (Phila Pa 1976) 2013; 38: pp. 2028-2037.
  • 20. Agus H, Kayali C, Arslantas M, et. al.: Nonoperative treatment of burst-type thoracolumbar vertebrafractures: clinical and radiological results of 29 patients. Eur Spine J 2005; 14: pp. 536-540.
  • 21. Dendrinos GK, Halikias JG, Asimakopoulos A: Factors influencing neurological recovery in burst thoracolumbar fractures. Acta Orthop Belg 1995; 61: pp. 226-234.
  • 22. Gertzbein SD, Court-Brown CM: Flexion-distraction injuries of the lumbar spine: mechanisms of injury and classification. Clin Orthop Relat Res 1988; 227: pp. 52-60.
  • 23. Cantor JB, Lebwohl NH, Garvey T, et. al.: Nonoperative management of stable thoracolumbar burst fractures with early ambulation and bracing. Spine (Phila Pa 1976) 1993; 18: pp. 971-976.
  • 24. Rechtine GR: Nonsurgical treatment of thoracic and lumbar fractures. Instr Course Lect 1999; 48: pp. 413-416.
  • 25. Radcliff K, Su BW, Kleper CK: Correlation of posterior ligamentous complex injury and neurological injury to loss of vertebral body height, kyphosis, and canal compromise. Spine (Phila Pa 1976) 2012; 1: pp. 1142-1150.
  • 26. Radcliff K, Klepler CK, Rubin TA, et. al.: Does the load-sharing classification predict ligamentous injury, neurological injury, and the need for surgery in patients with thoracolumbar burst fractures? Clinical article. J Neurosurg Spine 2012; 16: pp. 534-538. LOE I
  • 27. van Middendorp JJ, Patel AA, Schuetz M, et. al.: The precision, accuracy and validity of detecting posterior ligamentous complex injuries of the thoracic and lumbar spine: a critical appraisal of the literature. Eur Spine J 2013; 22: pp. 461-474. LOE III



  • Treatment of Thoracolumbar Burst Fractures



    Philippe Phan
    Polina Osler
    Kirkham Berwick Wood

    Anatomy, Classifications, and Radiologic Findings Related to Burst Fractures


    Anatomy of the Thoracolumbar Junction, Spinal Cord, and Cauda Equina


    The thoracolumbar junction is typically defined as extending from T10 through L2 (inclusive). This region accounts for more than half of all spinal fractures because it is located at the junction of the mobile lumbar lordosis and the more rigid thoracic kyphosis. Burst fractures make up 10% to 20% of all fractures in this transitional region. The purpose of this chapter is to review pathophysiology, classification, and treatment of thoracolumbar burst fractures.


    Unique to this region is also the transition from a spinal cord and the conus medullaris (upper motor neuron) to the cauda equina (lower motor neuron). The conus medullaris usually starts at T11 and ends around the L1 to L2 disc space. When it extends lower into the lumbar spine, it is often associated with a hypertrophic filum terminale and tethering of the spinal cord. In traumatic situations, the level of the injury and its localization with respect to the level of conus medullaris determines the degree of neurologic damage. As an example, a severe burst fracture at T12 might have elements of upper motor cord level injury (bowel and bladder dysfunction) and lower motor neuron pathology (lumbar radicular pain and weakness). The space available for the cord changes from its narrowest region in the midthoracic spine to an increased width in the lumbar region where the composition of the cauda equina made of peripheral nerves are more resilient to injuries.


    Burst Fractures in Thoracolumbar Fracture Classifications


    In 1943, Watson-Jones first described the comminuted wedge fracture, which would today be called a burst fracture, and highlighted the role of the integrity of the posterior ligamentous complex (PLC) in spinal stability. By the end of that same decade, Nicoll’s classification of thoracolumbar spine trauma further described the morphology of fractures based on a review of injuries sustained by coal miners. It defined stability using an anatomic classification based on four specific structures: the vertebral body; the disc; the intervertebral joints; and most important, the interspinous ligaments (ISLs). It was not until 1970 that Holdsworth. semantically distinguished “wedge compression fractures” from “compression burst fractures,” but in his classifications, both fracture types were stable.


    By the 1980s, the term burst fracture was defined in Denis’ classification as one of the four main categories of thoracolumbar injuries. His classification divides the spine into three columns with an emphasis on the middle column, which comprises the posterior vertebral body, posterior disc, and posterior longitudinal ligament. Mechanical instability was defined when two of the three columns were disrupted; therefore, all burst fractures were considered unstable according to this scheme. Denis’ classification also defined neurologic instability when a neurologic deficit occurred in the setting of a spinal fracture.


    The evolution in describing burst fractures is highlighted by the difficulty so many authors had in assessing its stability, which ultimately should guide treatment. Therefore, based on 100 computed tomography (CT) examinations of thoracolumbar injuries, McAfee and colleagues further classified them into six categories, including distinct categories for stable and unstable burst fractures. That classification emphasized the role of the PLC (including the facet joints) as a major factor in fracture stability.


    McCormack and colleagues have described the load-sharing classification for burst fractures. It is theorized and proven practically that the vertebral bodies’ ability to share load depends on the degree of comminution. In this system, the degree of comminution, fracture fragment displacement, and deformity are each graded 0 to 3 points and summed. Higher scores indicate greater instability and a greater likelihood of failure of short segment fixation. The system has been shown to have good interobserver and intraobserver reliability. In an effort to guide management, the Spine Trauma Study Group developed the Thoracolumbar Injury Classification and Severity Score (TLICS; Table 35B-1 ), which considers burst fractures as an intermediate severity fracture pattern based on radiographic appearance. Surgical recommendations for burst fracture are then guided by both neurologic status and the integrity of the PLC as many classifications have suggested in the past for burst fracture. That classification has demonstrated good intrarater and interrater reliability and validity and will therefore be the basis for the treatment of thoracolumbar fractures in this chapter.



    TABLE 35B-1

    THORACOLUMBAR INJURY CLASSIFICATION AND SEVERITY SCORE *




































































    Injury Characteristic Qualifier Score
    Injury Morphology
    Compression 1
    Burst +1
    Rotation and translation 3
    Distraction 4
    Neurologic Status
    Intact 0
    Nerve root 2
    Spinal cord, conus medullaris Incomplete 3
    Complete 2
    Cauda equina 3
    Posterior Ligamentous Complex Integrity
    Intact 0
    Suspected or indeterminate 2
    Disrupted 3

    PLC, Posterior ligamentous complex.

    * Score <4 degrees, nonsurgical; 4 degrees, nonsurgical or surgical; and >4, surgical.



    Radiologic Findings of the Burst Fractures


    Radiographic findings of a burst fracture on an anterior-posterior (AP) radiograph or coronal CT reconstruction are an increase in the interpedicular distance compared with the level above or below and a decrease in vertebral height. On the lateral radiograph or sagittal CT scan, findings can include a decrease in vertebral height, lack of definition of the contour of the cortices of the vertebral body with retropulsion of the posterior cortex into the spinal canal, local kyphosis around the fractured vertebra, and an increase in the interspinous distance ( Fig. 35B-1, A ). An increase in the interspinous distance should increase the level of suspicion for disruption of the PLC. Comparing this interspinous distance between supine and upright images may identify a functionally incompetent PLC. CT scans are used to further detail the fracture pattern and assess instability (widening of the facet joint) and canal compromise at the level of injury ( Fig. 35B-1, B ).




    Figure 35B-1


    Lateral ( A ) and anterior-posterior ( B ) plain radiograph views of the lumbar spine showing an L1 burst fracture (small arrow) with retropulsion of bone into the spinal canal (large arrow).


    The Posterior Ligamentous Complex


    As originally defined by Holdsworth, the PLC is generally defined by five components at the level involved: the supraspinous ligament (SSL) and ISL, the ligamentum flavum (LF), and the right and left facet capsules. Vacarro and colleagues. also added the thoracolumbar fascia as part of the PLC. Determination of PLC integrity, however, remains a topic of controversy despite its central role in management determination.


    Physical examination may help in the assessment of the integrity of the PLC. Severity of pain, on palpation of the spinous processes, presence of hematoma, or identification of gaps between them are indicators of disruption of the PLC. Further simple logrolling and assessing pain response can help measure stability of the fracture.


    Radiologic imaging has proven helpful but, unfortunately, not definitive in all cases. Vacarro and members for the Spine Trauma Study Group extracted 12 criteria considered important to evaluate PLC disruption on magnetic resonance imaging (MRI) from the published English literature since 1949. Those criteria were posterior edema on short tau inversion recovery (STIR) MRIs; disrupted PLC components (i.e., LF, ISL, or SSL) on MRI; focal kyphosis without vertebral body injury; interspinous spacing greater than that of the level above or below on an AP plain radiograph; palpable interspinous defect on examination; focal posterior tenderness on examination; diastasis of the facet joints on radiograph, MRI, or CT; avulsion fracture off the superior or inferior aspect of contiguous spinous processes; history of the injury mechanism; more than 50% compression of the anterior vertebral body on lateral plain radiographs without fracture of the posterior wall on CT scan; and vertebral translation ( Fig. 35B-2 ). Twenty-eight surgeons returned the survey, and more than half ranked “vertebral body translation” as the most important factor followed by increased interspinous spacing and diastasis of the facet. Although Denis and the Arbeitsgemeinschaft für Osteosynthesefragen (AO) classification rely mainly on bony morphology from radiographs and CT scans, the emphasis on the role of the PLC to spinal instability has led to the inclusion of MRI as part of spinal fracture evaluation. MRI was often considered a critical modality to evaluate PLC integrity. In that study, disrupted PLC components (i.e., LF, ISL, or SSL; “black stripe”) on T1-weighted sagittal MRI was the fourth most important factor in evaluating PLC disruption. However, MRI findings should be considered with caution. Vaccaro and colleagues have highlighted low-specificity values for detecting ISL and SSL injuries on MRI, suggesting a large proportion of false-positive injuries on MRI compared to preoperative surgical findings. Thus, despite the central role of PLC integrity in the evaluation and decision making for thoracolumbar spine injuries as suggested by the TLICS, there remain few strict criteria or clear definition in defining a PLC injury, and the exact prognosis associated with PLC injuries related to radiographic, neurologic, and patient-reported outcomes has yet to be defined.




    Figure 35B-2


    Sagittal ( A ) and axial ( B ) T2-weighted magnetic resonance imaging views of the lumbar spine showing the L1 burst fracture (white arrow) with compression of the spinal cord at the same level with surrounding regions of increased signal intensity corresponding to the surrounding edema (black arrows). The degree of the canal narrowing can readily be seen on the axial image on the right .


    Treatment


    Nonoperative Treatment


    Indication


    In a stable burst fracture (without significant PLC disruption) and absence of a neurologic deficit, nonoperative treatment is the treatment of choice. In a randomized controlled trial (RCT) of 47 consecutive patients with stable burst fractures without neurologic deficit, Wood and colleagues found no major long-term advantage when comparing operative treatment with nonoperative treatment based on radiographic criteria, Short Form 36 (SF-36), and Owestry questionnaires, but more frequent complications and greater cost were found in the operative group. Those findings were not reproduced in a later RCT with 34 patients; Siebenga and colleagues found that for stable burst fracture without neurologic deficit, short segment stabilization provided less kyphotic deformity, better functional outcomes, and higher return to original work than in the nonoperative group. They concluded that in addition to fracture stability and neurologic findings, treatment should also take into consideration the amount of deformity and patient preference with respect to complication risks ( Fig. 35B-3 ).




    Figure 35B-3


    Lateral and anterior-posterior (AP) plain radiographs of the thoracolumbar spine of a patient with an L1 burst fracture, which was treated nonoperatively with a thoracolumbar orthosis. A, Lateral plain radiograph made with patient in a thoracolumbar orthosis at the time of discharge from the hospital, demonstrating a 22-degree kyphosis. B and C, AP and lateral views, respectively, of the same patient taken 36 months after the injury, demonstrating progression of the deformity to 28 degrees of local kyphosis.

    (Figure reproduced with permission from Wood K, Buttermann G, Butterman G, et al: Operative compared with nonoperative treatment of a thoracolumbar burst fracture without neurological deficit. A prospective, randomized study, J Bone Joint Surg Am 85(5):773–781, 2003.)


    Bracing


    Traditionally, stable thoracolumbar burst fractures have been treated with a custom thoracolumbar spinal orthosis (TLSO) or a hyperextension brace such as a Jewett brace. Thoracic burst fractures from T5 to T10 may be treated similarly. Upper thoracic burst fractures from T1 to T5 are treated with a TLSO with neck extension.


    The use of bracing may not always be necessary, however. Two recent RCTs compared bracing with no bracing in patients with stable burst fractures at the thoracolumbar junction. These studies found that bracing had no effect on pain, radiologic results, or functional outcome. In our opinion, bracing is still considered for lower lumbar fractures (rigid lumbosacral orthosis [LSO]), stable thoracolumbar fractures with intractable pain or concern for patient compliance (rigid TLSO), and a lower thoracic fracture at risk of kyphosis (orthosis with upper trunk support such as a Jewett brace or a rigid TLSO with neck extension).


    Activity


    When the patient is able to stand, an upright radiograph should be taken to ensure that upright loading at the fracture site does not produce an increased deformity. When conservative treatment is chosen, ambulation should be initiated as early as possible to avoid complications associated with bed rest (deep venous thrombosis leading to pulmonary embolism, pneumonia, or decubitus ulcers). In the first 3 to 6 months, activity with risks (manual labor, prolonged seating, impact sports, bending or heavy lifting) should be avoided. A progressive increase in activity with the above restrictions is allowed with close follow-up at 1, 4, 8, and 12 weeks. Patients are warned to seek attention if there is the development of weakness or numbness or changes in bowel or bladder function. During follow-up visits, close attention is paid to the deformity at the fracture site. If there is an increase in the deformity (kyphosis >30 degrees or scoliosis >10 degrees), intractable pain or neurologic symptoms, surgery may be considered. When there is radiographic evidence of fracture healing, flexion and extension radiographs are taken to exclude motion at the fracture site.


    Geriatric Considerations


    Senile fractures from osteoporosis in elderly patients should also be closely monitored with frequent radiologic examinations in the first month because they are at higher risk of kyphosis deformity leading to neurologic compromise. Patients and family members should be warned to watch neurologic function closely. Assessment of osteoporosis is recommended to limit further fractures (particularly the spine, hip, wrist, and shoulder) and pharmacologic treatment is implemented if needed.


    Surgical Treatment of Burst Fractures


    Surgical Indication


    Surgical treatment of thoracolumbar burst fractures is indicated when there is evidence or risk of neurologic worsening or stability compromise. For this reason, the TLICS uses neurologic status and two factors influencing stability (injury morphology and PLC integrity) as components influencing surgical decision. The approach and surgical technique to use should take into consideration three main criteria:



    • 1.

      The fracture pattern


    • 2.

      The ability to decompress neural elements (based on fracture pattern and age of the fracture)


    • 3.

      The presence of PLC lesions



    Other indications for surgical treatment are when associated with multiple injuries or in those who cannot be managed with an orthosis such as very obese individuals.


    Burst fractures can be treated with an anterior, posterior, or combined approach for decompression and instrumentation. Posterior spinal instrumentation at the thoracolumbar junction is well established and has its advantages, including familiarity of the posterior midline approach; the lack of pulmonary, visceral, and vascular structures; and the ability to relatively safely reexplore the surgical site if necessary. In most acute cases, decompression and stabilization can be achieved from a posterior approach with limited associated comorbidities ( Fig. 35B-4 ). Posterior instrumentation also permits stabilization of the posterior elements when the PLC is torn. Limitations of the posterior approach are that the decompression is indirect (usually achieved from reduction of bony fragments by tension of the posterior longitudinal ligament and ligamentotaxis) or, alternatively, by the need to mobilize the dural sac to achieve proper decompression. Further hardware failure and reoperations are common, especially when the vertebral bodies have more severe injury and comminution.




    Figure 35B-4


    Sagittal ( A ) and axial ( B ) views of the T2-weighted magnetic resonance image of a 51-year-old woman who fell after completing a ski jump and sustained a severe burst fracture of L1 with disruption of both endplates, 20 degrees of local kyphosis, and 65% canal compromise caused by bony retropulsion. She was neurologically intact. Postoperative anterior-posterior ( C ) and lateral ( D ) radiographs show restoration of sagittal alignment with fixation and posterolateral fusion from T11 to L3.


    In the past, fusion was routinely performed in conjunction with posterior instrumentation for burst fractures. Fusion-less fracture care has the advantage of minimizing stiffening of spinal segments while still providing stabilizing forces while the fracture heals. This can be performed using an open or minimally invasive technique. In most cases, hardware is removed 9 to 12 months later. Dai and colleagues reported the results of a randomized trial comparing fusion with no fusion after instrumentation for thoracolumbar burst fractures. At 5 to 7 years of follow-up, they found no difference in clinical and radiologic outcomes between the two groups. Tian and colleagues performed a meta-analysis of four studies comparing fusion with no fusion. The fusion-less group had shorter operative times, but no difference in clinical results or complications were present between groups. Radiologic results favored the fusion-less group that had less postoperative vertebral height loss. Complications and hardware failure were not different between groups. The topic is discussed further in Chapter 35E .


    Anterior reconstruction with instrumentation can overcome some of the limitations of the posterior approach when direct visual decompression can be achieved, particularly in the case of subacute or chronic fracture in which the bony fragments can only be mobilized with difficulty. Also, the anterior approach allows application of a large distractive force to be applied, restoration of the mechanical integrity of the anterior column, removal of torn or damaged discs, direct decompression of the spinal canal, potential avoidance of iliac crest harvesting, and frequently fusion of fewer levels of the spine.


    The combined anterior approach is often indicated for the more severe fractures with neurologic deficits. If a posterior fusion did not result in complete decompression then an anterior decompression should be considered. Similarly, if adequate stability was not achieved after posterior instrumentation and patients are at risk or are developing kyphosis, then anterior reconstruction should be considered. For highly unstable burst fractures treated initially anteriorly, posterior instrumentation should be considered.


    In a neurologically intact patient with burst fractures, anterior or posterior approaches have proven to be equivalent for radiographic and patient-reported functional outcomes in a series by Wood and colleagues. Similar results were found in an RCT from Lin and colleagues for burst fractures with neurologic deficit in which radiologic and motor score improvement did not show any significant difference between the two approaches. However, different rates of complications were reported in the two RCTs. Wood and colleagues found that the anterior approach had fewer complications and reoperation; in their series, 17 complications were seen in their posterior group with hardware removal of painful instrumentation in nearly one-third of those patients. To the contrary, Lin and colleagues have suggested that posterior approach would lead to less blood loss, shorter operative times, and better pulmonary function postoperatively.


    Currently, there is limited consensus about the effect of timing of surgery on neurologic recovery. Clinical judgment should be taken into consideration, and, in general, spinal care should be handled after the treatment of life- and limb-threatening injuries.


    The Anterior Thoracolumbar Approach with Corpectomy and Instrumentation


    Surgical Indication, Advantages, and Limitations.


    The anterior thoracolumbar retroperitoneal approach is most commonly indicated for burst fractures between T10 and L3, which can also benefit from direct anterior decompression. Other indications include a large retropulsed fragment with marked canal compromise (>50%) and incomplete neurologic deficit and anterior comminution and kyphosis less than 30 degrees requiring additional anterior column support. After a posterior stabilization, an anterior approach should be considered when there is suboptimal neural recovery resulting from inadequate canal ( Fig. 35B-5 ).




    Figure 35B-5


    Anterior-posterior (AP) ( A ) and lateral ( B ) plain radiographs of a 54-year-old woman who was involved in a roll-over motor vehicle accident with a sustained burst fracture at L1. Axial computed tomography (CT) reconstruction ( C ) shows the appearance of the burst fracture with bone retropulsion resulting in 55% canal occlusion. The bottom row shows the postoperative appearance of the injury: AP ( D ) and lateral ( E ) plain radiographs of the lumbar spine taken 2 years after the surgical stabilization show maintenance of correction. Axial CT reconstruction ( F ) at 2 years after surgery shows only modest 5% residual canal occlusion.

    (Reproduced with permission from Wood KB, Bohn D, Mehbod A: Anterior versus posterior treatment of stable thoracolumbar burst fractures without neurologic deficit: a prospective, randomized study, J Spinal Disord Tech 18(suppl):S15–S23, 2005, Fig. 1, A-F.)


    The main advantages of the anterior thoracolumbar approach are the possibility to directly visualize the fracture, directly decompress the spinal cord from the side of the pathology, and reconstruct the anterior column. This reconstruction is thought to allow a better correction of the kyphotic deformity than posterior approach. The main limitation of this approach is that it only addresses the ventral or compression side of the spine. Therefore, an additional posterior approach may be required for stabilization in cases in which the PLC and its tension band effect are lost.


    For burst fractures of T10 and above, anterior decompression and fusion are performed using a transthoracic approach usually from the right side. A video-assisted transthoracic (VATS) approach may minimize pulmonary injury but is technically demanding and application of instrumentation is challenging.


    Surgical Technique: Anterior


    Positioning.


    For fractures below T12, a left-sided approach with the patient positioned in a right lateral decubitus position is preferred to avoid retraction of the liver and damage to the frail vena cava. Patients are positioned in a lateral decubitus position ( Fig. 35B-6 ). After the patient is positioned, attention is given to check all pressure points, particularly the brachial plexus, under an axillary roll. The elbows and knees are protected with gel pads to protect the ulnar and common peroneal nerves, respectively. The patient might be positioned with the thorax and abdomen leaning slightly forward to allow abdominal contents to fall forward during the exposure. Some surgeons prefer straight lateral positioning to maintain orientation of the vertebral body during the instrumentation. To improve exposure, the fracture level can be placed over the table break and, depending on fracture stability and neurologic involvement, the table can be broken to raise the vertebra of concern into the operative field. At the time of fixation, it is necessary to level the table to avoid the introduction of scoliosis deformity.














    Figure 35B-6


    A, Positioning of the patient for the anterolateral approach. The arrow points toward the head of the patient. The patient is positioned on the right side, and the left side is exposed and prepared for the approach. This allows the surgeon to avoid damaging the liver during the approach to the spine. Technique of anterior transthoracic corpectomy and fusion. B, The patient is placed in a straight decubitus position with the shoulders extended forward 90 degrees, neutral in terms of abduction and adduction, and with the elbows straight. Care is taken to protect the downside brachial plexus by using a pad just distal to the axilla. The dotted line over the rib represents the incision one level above that of the spinal fracture. C, If the incision is used to expose above the T6 rib, the posterior limb of the incision is extended cephalad halfway between the medial border of the scapula and the spinous processes. All of the intervening muscles down to the chest wall are divided and tagged for later repair. D, After the thoracic cavity has been entered, the self-retaining chest retractor is inserted. The parietal pleura is incised halfway between the anterior great vessels and the posterior neural foramina, and the segmental vessels are ligated at this same level. The vertebra to be excised as well as one vertebra above and one vertebra below are exposed. Extraperiosteal dissection provides the best plane. A malleable retractor is placed on the opposite side of the spine and connected to the self-retaining chest retractor with a clamp. This malleable retractor serves to protect the great vessels during the vertebral corpectomy. E, A scalpel and rongeur are used to remove the discs above and below the level of the vertebral fracture. F, An osteotome, chisel, or gouge is used to excise the vertebral body back to its posterior cortex. Special care is taken to originally position the patient exactly in the straight decubitus position. During the vertebral body resection, using these instruments, each of the cuts is made perpendicular to the floor. These instruments can be used as long as red cancellous bone is encountered. As soon as white cortical bone is encountered, these instruments should no longer be used. G, A high-speed burr can be used to perforate the posterior vertebral body cortex into the spinal canal. When the neural compression is significant, a diamond-tipped burr can be used to minimize the chances of dural or neural injury. H, Down-biting 90-degree Kerrison rongeurs are used to remove the bone on the most superficial portion of the vertebral body. I, Reverse-angle curettes are used to carefully impact the bone from the spinal canal on the far side of the vertebral body. J, The bone resection at the end of the decompression should extend from the pedicle on one side to the pedicle on the opposite side. It is easy to underestimate the extent of bone removal necessary to achieve this. At the end of the neural decompression, the dura should bulge anteriorly in a uniform fashion from the endplate of the vertebra above to the endplate of the vertebra below and from pedicle to pedicle. If the dura does not bulge out concentrically, the surgeon should check for residual neural compression. K, After the corpectomy and resection of the disc above and below the corpectomy. If there is any degree of osteoporosis present, the trough should be cut through the cancellous bone up to the next intact endplate at the superior end of the cephalad vertebra and the inferior aspect of the caudal vertebra. A ridge of bone should be preserved at the posterior aspect of these adjacent vertebrae to prevent migration of the bone graft into the spinal canal. L, At the end of the neural decompression and fusion, there should be adequate space between the bone graft and the dura and neural elements to minimize the chance of producing any iatrogenic neural compression. This illustration shows three strips of rib being used as bone graft, but a single large piece of iliac crest can also be used and may actually provide a stronger anterior strut. A transverse section at the vertebrae above ( M ) and below ( N ) the level of the corpectomy should reveal an adequate posterior rim of cortical bone to prevent migration of the bone graft into the spinal canal and an anterior cortical and cancellous rim of bone to prevent dislodgement of the bone graft.


    For lower thoracic injuries, a transthoracic approach is used. Intubation is performed with a double-lumen tube that allows deflation of the right lung during surgery. The patient is positioned in the left lateral decubitus position with the right side up ( Fig. 35B-6, B ). Similar precautions to protect neurovascular structures are carried out as described earlier.


    Landmark, Incision, and Approach.


    The incision is based on the vertebra affected. Given the cephalocaudal direction of the ribs, the incision should be made on the rib one to two levels above the region of interest. For example, to treat a L1 burst fracture, the incision should be centered on the 11th or 12th rib. The incision follows the ribs on its whole length. After skin incision, the abdominal muscle layers are transected in line with the rib in the following order from posterior to anterior: latissimus dorsi and external oblique; serratus posterior inferior and internal oblique, transversus abdominis, sacrospinalis, and multifidus.


    The rib is then dissected subperiosteally, and particular care is given to use an elevator circumferentially, particularly to avoid damage to the subcostal neurovascular bundle. The rib is cut anteriorly at the cartilaginous junction and posteriorly close to the rib head. A self-retaining retractor is then placed. When the 12th rib is excised, its tip is at the junction of the transversalis fascia, pleura, and diaphragm, which facilitates the identification of the retroperitoneal space and dissection of the pleura, which is reflected upward while the quadratus lumborum is retracted downward. Above the 12th rib, the pleural space is accessed through a small incision, which is extended in line with the incision while protecting the underlying structures.


    When necessary, the diaphragm is detached from its insertion above the arcuate ligament, leaving 2 cm of the periphery to aid repair at closure. To facilitate closure, it is recommended to leave suture tags of different colors along the diaphragm incision.


    In the retroperitoneal space, the peritoneum, retroperitoneal fat, and kidneys are reflected anteriorly to expose the quadratus lumborum and the psoas muscle overlying the vertebral bodies. A malleable retractor can be used to protect and maintain the viscera anteriorly. The psoas is dissected from anterior to posterior of the lateral aspect of the vertebral body until the base of the pedicle is palpated and the ventral neural foramen exposed. In the case of a transthoracic decompression, after the rib is removed, the pleura is opened, and the lateral aspect of the vertebral bodies is exposed. The right lung can then be deflated.


    The fractured vertebral body is then localized based on the surrounding hematoma and the deformity and confirmed by an intraoperative radiograph ( Fig. 35B-6, C ). Upon confirmation of the adequate level, the fractured and adjacent levels are exposed by releasing the overlying pleura followed by isolation and ligation of the segmental arteries, which allows the great vessels to be mobilized and facilitates a safe approach to the contralateral side if needed.


    Corpectomy and Decompression.


    After adequate exposure is completed, discectomies are performed between the vertebra of interest and the adjacent levels ( Fig. 35B-6, D ). The corpectomy can be started using the pedicle as a guide for the posterior margin and the depth of the discectomy for the contralateral border. Using osteotomes or chisels, the middle and anterior vertebral body is resected about 1 cm anterior to the pedicle base ( Fig. 35B-6, E ). The bone from the corpectomy is kept for bone grafting. After an adequate corpectomy is achieved to the contralateral wall and pedicle, decompression of the dural space is performed. The posterior wall is first thinned using a high-speed diamond burr and curettes and Kerrison rongeurs then complete the decompression which should span between the disc spaces and from ipsilateral pedicles to contralateral pedicles ( Figs. 35B-6, F to H ).


    Instrumentation and Fixation.


    After the corpectomy, the overall alignment is checked and reduction maneuvers applied as needed with manipulation of the spine from external pressure or distraction within the corpectomy site applied with a laminar spreader or using an expandable cage mechanism. Alternatively, the instrumentation can be used as sites for distractors to regain lordosis and vertebral height ( Fig. 35B-7, A and B ). Tricorticate iliac crest, allograft femurs or humeri, titanium mesh cage, and expandable cages are possible options to provide anterior support or fusion. When using cages, vertebral corpectomy bone graft is placed in the cage while structural bone graft from rib resections can be used to bridge both endplates longitudinally. Superior and inferior endplates are prepared using a curette or a burr to promote bleeding and vascular supply to the graft, but attention should be taken not to compromise the mechanical support of the endplates, which can lead to graft impaction and kyphosis. Additional support with internal fixation can be provided with transvertebral screws and rods or plates ( Fig. 35B-7, C to F ). Screws are placed laterally toward the contralateral side with care to not injure the visceral and vascular structure. The endplates and vertebrae posterior walls are used as guides for screw orientation. Precise measurement of the screw length when the opposite cortex is breached with the ball-tip probe is essential. Upon placement of one or two screws in each of the adjacent vertebrae, one or two rods or a plate is placed and gentle compression can be applied to secure the anterior graft or cage in place.




    Figure 35B-7


    Technique for anterior spinal instrumentation after corpectomy. A, After using a depth gauge directed on the exposed vertebral body, appropriately sized screw lengths are selected to engage the opposite cortex of the vertebral body. The bolts are placed parallel to the adjacent endplate to avoid intrusion into the disc space above and below the corpectomy site. B, Distraction is applied against the bolts, allowing easy insertion of the strut graft into the corpectomy site. C, Determination of proper length of plate via a template is important to avoid impingement of the superior or inferior disc space. Locking nuts are applied and provisionally tightened. D, Compressive forces are applied and locking nuts are tightened firmly. E and F, Finally, two anterior screws are placed, and the nuts are crimped down, preventing possible backing out or loosening.

    (Redrawn with permission from Zdeblick TA: Z-Plate-ATL Anterior Fixation System: Surgical Technique. Sofamor Danek Group, Inc. All rights reserved.)


    Closure and Postoperative Course.


    Before closure, thorough hemostasis is ensured and, if possible, the pleura is repaired over the hardware. A chest tube is placed under direct visualization and the diaphragm os closed following the suture stitches left upon opening. The abdominal musculature is closed in multiple layers.


    Postoperative clinical and radiographic monitoring of the chest guide the timing of the removal of the chest tube when there is no evidence of pneumothorax and when drainage has subsided. Postoperative ileus is common and a nasogastric tube is placed as needed.


    Postoperatively, a thoracolumbar orthosis may be used up to 12 weeks. Its use is at the surgeon’s discretion depending on the level of instability, bone quality, the addition of posterior fixation, and patient compliance. Standing radiographs of the thoracolumbar spine are taken when the patient is able to stand to ensure construct integrity under physiologic loads. Ambulation is initiated postoperatively as early as possible, but physical therapy typically begins months after surgery when a successful fusion seems likely.


    Complications from Anterior Thoracolumbar Approach.


    Complications can be grouped into three cate­gories, surgical approach, decompression, and structural integrity.


    Early complications related to the approach include pneumothorax, atelectasis, pneumonia, ileus, infection, nerve injury (particularly genitofemoral nerve and nerve root from the lumbar plexus), and visceral and vascular lesions. During decompression, particular care must be taken not to injure the dural sac. If a spinal fluid leak is noted from iatrogenic manipulation or secondary to the trauma, early repair should be attempted. If this fails, cerebrospinal fluid (CSF) drainage via a lumbar drain is recommended. In thoracic cases, the negative pleural pressure keeps CSF leaks from sealing; therefore, lumbar drainage should be considered.


    Late complications include incisional hernia. Other late complications are related to disruption of the structural integrity of the construct from bone and bone–instrument failure or pseudarthrosis, which have both been described between 5% and 10%. The frequency of those complications has lessened over the years with the evolution of modern rigid spinal instrumentation, including addition of fixation to the corpectomy graft.


    Surgical Technique: Posterior


    The technique of posterior instrumentation for burst fractures is described in Chapter 35C .


    Outcome


    Nonoperative Treatment versus Operative Treatment for Thoracolumbar Fractures without Neurologic Deficit


    In the thoracolumbar burst fractures without neurologic deficit, a meta-analysis from Gnanenthiran and colleagues retrieved four major trials for a total of 79 randomized patients (41 with operative treatment and 38 with nonoperative) with a mean follow-up time from 24 to 118 months. No differences in pain, Roland Morris Disability Questionnaires scores, or return to work rates were present between operative and nonoperative groups. At approximately 4 years, the operative group had an improvement in kyphosis from baseline by 1.8 degrees with a mean of 11 degrees, as opposed to a kyphosis of 16 degrees with a worsening of 3.3 degrees in the nonoperative group. Although the amount of correction was significantly different between the treatments, the final degree of kyphosis was not. Operative treatment was also associated with higher complication rates and costs. There was also no association found between kyphosis and pain. Return to work rate were 67% for the nonoperative group compared with 70% in the operative group, a statistically insignificant difference.


    Comparison of Anterior and Posterior Approach


    Wood and colleagues conducted a prospective randomized study comparing anterior versus posterior treatment of stable thoracolumbar burst fractures without neurologic deficit or loss of structural integrity of the PLC. Thirty-eight patients had a minimum 2-year follow-up, 18 had posterior spine fusion and 20 had an anterior approach. Although lengths of hospital stay and operating times were similar, blood loss was higher in the anterior group, but significantly more adverse events (instrumentation removal, wound dehiscence, instrumentation, or bone failure) were observed in the posterior group. Clinical results between the two groups were similar and demonstrated good pain reduction based on visual analog scale (VAS) scores, improvement in disability score (RMFDS) to 8.3 for posterior group and 8.9 for the anterior group, and ability to return to work after 6 months of 44% for the posterior group and 55% for the anterior group. The study concluded that anterior surgery might lower complications and prevent additional surgeries.


    Lin and colleagues also conducted a prospective randomized study comparing anterior versus posterior treatment of thoracolumbar burst fractures but included patients with neurologic deficits and unstable fracture patterns. A total of 64 patients were recruited and equally distributed in comparable group for baseline characteristics. At a minimum of 2-year follow-up, all patients achieved adequate solid fusion and significant neurologic improvement independent from the approach used. In that study, it was found that the posterior approach was associated with less blood loss, complications, shorter operative time, and better postoperative pulmonary function. Although it is often believed that an anterior approach provides a better decompression from direct visualization, this study demonstrated that the posterior approach provided as good a decompression and neurologic improvement while having fewer perioperative complications.


    Comparison of Posterior Instrumentation with and without Fusion


    Fusion-less fracture care provides opportunity for stabilization and reduction without permanent stiffening of the spine. Dai and colleagues reported results of RCTs at 5 to 7 years of follow-up in 73 patients who had stable-type vertebral burst fractures. Patients were treated by short segment instrumentation with or without fusion. Operating time and blood loss were less in the fusion-less group. Clinical and radiologic results did not differ between groups. No differences in complications were present, although donor site pain was present in two-thirds of the fused patients.


    In a meta-analysis of four studies including 220 patients, Tian and colleagues compared fusion to fusion-less instrumentation for thoracolumbar burst fractures. Similar to the results by Dai, no difference at follow-up was present between groups. Operating time and intraoperative blood loss were less in the fusion-less group. No differences were present in rates of hardware failure or surgical complications.


    Conclusions


    Thoracolumbar burst fractures are common injuries with the hallmark being a fragment from the posterior vertebral wall that is retropulsed into the spinal canal. Two important factors drive treatment decisions: the neurologic status and the integrity of the PLC. Treatment decisions are now based on a new classification scheme that takes these two variables into account. Stable-type fractures are best treated nonoperatively. Patients with neurologic deficits can be treated by either anterior or posterior approaches or even combined approaches with the goal of decompression, reduction, and stabilization. For other unstable fractures, posterior instrumentation should be performed. When posterior instrumentation is used, a fusion-less technique may be appropriate, but this requires further investigation.




    Identification, Classification, Mechanism, and Treatment of Thoracolumbar Fracture-Dislocations



    John R. Dimar
    Paul C. Celestre
    Ashish Upadhyay
    Nandita Das

    Thoracic and upper lumbar spine injuries are among some of the most common types of spinal injuries. Fracture-dislocations of the thoracolumbar spine have long been recognized as potentially unstable spine injuries that often present with significant neurologic injury. These injuries are generally a result of high-energy trauma, such as high-speed motor vehicle accidents, falls from heights, and occupational injuries, and are frequently associated with polytrauma. Caring for patients who have undergone such traumatic spinal fractures is challenging because a majority of these patients also have concomitant polytrauma with injury to other organ systems. Commonly associated injuries that occur with fracture-dislocations include closed head injuries, pulmonary contusions, splenic and hepatic lacerations, bowel injuries, vascular injuries, long bone fractures, and neurologic injuries. Because spinal fracture-dislocations have the potential to be highly unstable, these injuries require diligence in providing safe transport to the emergency department; rapid and precise diagnosis; and, when the patients are medically stable, emergent surgical reduction, decompression, and primary stabilization to prevent secondary neurologic injury. Flexion-distraction injuries, commonly referred to as Chance fractures, are a unique subset of high-energy thoracolumbar spine injuries. These injuries commonly result from rapid deceleration during motor vehicle accidents. Chance injuries are unstable and commonly require surgical stabilization.


    Proper determination of the mechanism of injury is very important for appropriate surgical management of the patient. To expedite the treatment of traumatic injuries to the spine, various spinal injury classification systems have evolved over time, and each generation of spinal trauma surgeons has refined and modified the systems to describe the most commonly recognized fracture patterns.


    Classification


    Fracture-Dislocations


    Fracture-dislocations of the thoracolumbar spine result from high-energy injuries. Most commonly, the rostral vertebra is translated anteriorly on the caudal vertebra, but it is not unusual to see injury patterns that include retrolisthesis or lateral translation. Although they can be classified by direction of dislocation, this is arbitrary and practically makes little difference in regard to treatment. They are devastating spinal injuries that are frequently accompanied by a complete neurologic deficit. Fracture-dislocations of the spine inherently disrupt all three columns and are thus unstable and treated with surgery.


    Flexion-Distraction Injuries


    Two definitions of flexion-distraction injuries have been recently proposed and evaluated for reliability.


    Definition A: Flexion-distraction injuries involve an axis of rotation anterior to the anterior longitudinal ligament as opposed to flexion-compression injuries that involve an axis of rotation within the vertebral body


    Definition B: Flexion-distraction injuries are any injuries in which there is disruption of the posterior ligamentous complex (PLC) but no retropulsion of the vertebral body into the canal (i.e., there can be some vertebral compression as long as it is not associated with retropulsion, which would make it a flexion-compression injury).


    In the study survey, 67% of the treating physicians agreed to definition B, and the remaining 33% agreed to definition A. In reality, these are similar injury patterns that both include disruption of the PLC. The more recent classification systems strongly suggest that these unstable injuries require immobilization or an open reduction, instrumentation, and fusion to restore alignment and prevent secondary neurologic injury. Here it is important to understand that the mechanism of injury will frequently guide the surgical planning. These definitions may be inadequate and truncated because these injuries have long been postulated to represent a continuum of a flexion and distraction moment where the PLC fails followed by the disruption of the facets, posterior longitudinal ligament, the disc, anterior longitudinal ligament, and finally displacement of the superior vertebral body on top of the other that often injures the spinal cord or cauda equina. Additionally, other forces can be part of the injury, including rotation, coronal angulation, extension, and compression, with the latter causing concurrent fracturing of the body.


    Another pattern similar to flexion-distraction injuries is the eponymous Chance fracture. In this injury, an extreme distraction moment imparted to the thoracolumbar spine results in the fracture line extending from the posterior spinouts process through the pedicles, vertebral body, and anterior longitudinal ligament. The injury usually involves primarily bone, but varying amounts of posterior ligamentous and discoligamentous injury may be present.


    Assessment of Severity of Injury


    White and Panjabi described the instability of the spine as follows: “Inability of the spine, under physiologic loads, to maintain the relationships between the vertebrae so that there is neither initial nor subsequent neurological deficit, deformity and/or pain.” In the clinical setting, this is hard to apply and therefore other methods to quantify the severity of injury have been developed.


    Vaccaro and the Spine Trauma Study Group (STSG) combined the morphologic parameter of the spinal injury with the integrity of the PLC and the degree of neurologic injury into a novel classification system known as the Thoracolumbar Injury Classification and Severity (TLICS) system. Because the integrity of the PLC, which is poorly visualized on computed tomography (CT) scanning, plays a key role in fracture evaluation using the TLICS system, the authors recommended the use of magnetic resonance imaging (MRI) studies to evaluate the integrity of the PLC. The goal of this novel classification system is to predict which fractures are unstable and require surgical stabilization. The system requires the surgeon to grade the fracture and assign points by evaluating three key components of the injury: neurologic status, fracture morphology, and integrity of the PLC ( Fig. 35C-1 ).




    Figure 35C-1


    A, Schematic Vaccaro Thoracolumbar Injury Classification and Severity (TLICS) three-part spine fracture classification system, which includes the mechanism of injury, the neurologic status, and the status of the posterior ligamentous complex. The total point tally guides the treatment decision. This is the first system to incorporate the neurologic status to determine surgical indications. B, TLICS classification flow chart.

    (From Vaccaro A, Zeiller SC, Hulbert RJ, et al: The thoracolumbar injury severity score: a proposed treatment algorithm, J Spinal Disord Tech 18:209–215, 2005, Figs. 2, 3, and 4.)


    At least two or three additional points are assigned to the total tally for any patient having a partial spinal cord injury or with cauda equina syndrome. This is then combined with the fracture morphology points, and the points are determined by evaluating the status of the PLC. Based on this system, any fracture scoring 5 points or above requires consideration for surgery, score of 4 is equivocal, and scores of 3 or less should receive nonoperative treatment. In the case of grading fracture-dislocations, which are inherently unstable, the system reliably grades the injury with a minimal score of 5 to 6 points and generates a surgical recommendation. The system is quick and intuitive to use and can provide a useful algorithm to formulate a potential treatment plan. The TLICS classification scheme assigns 7 points to Chance fractures, not including any points for neurologic deficits: 4 for pattern of injury and 3 for disruption of the PLC; thus according to the TLICS system, all Chance fractures meet operative criteria.


    The TLICS has been evaluated and shown to have good to excellent interobserver reliability. Owing to its intuitive classification algorithm and ease of use, it has become commonly used among current spinal trauma surgeons.


    Mechanisms of Injury


    Studies have shown that the direction of the force applied to the spine will directly determine the fracture pattern and consequently impact the level of stability. The focus of this discussion on the mechanism of injury is limited to flexion, flexion rotation, lateral shear, extension, and distraction forces, which, when severe enough, have the potential to disrupt the structural integrity of the spine, causing a fracture-dislocation. Thoracolumbar fracture-dislocations are generally thought to result from the acute transition of the rigid thoracic spine to the more mobile lumbar spine. White and Panjabi reported that in a cadaveric evaluation of the spine, the flexion and extension ranges of motion in the lower thoracic spine increase from 5 to 12 degrees from T6 to T7 to T12 to L1 compared with the lumbar spine, which averages 15 degrees. They also demonstrated that the axial rotation and lateral side bending of the thoracic spine is greater than that of lumbar spine because of the coronal orientation of the facets compared with the more sagittal lumbar facet joints.


    This mismatch in the range of motion and the rigidity of the thoracic spine secondary to rib buttressing compared with the lumbar spine makes the thoracolumbar junction susceptible to fractures when a severe and violent flexion, extension, rotational, or shear force occurs during high-energy trauma. This ultimately results in the force of the injury being concentrated at the thoracolumbar junction, producing a variety of fracture patterns that can render the spine unstable and subject the spinal cord and cauda equina to risk of injury. As a result of these biomechanical effects, almost 50% of all vertebral body fractures and 40% of all spinal cord injuries are concentrated at the thoracolumbar junction. The mechanisms that can create an unstable thoracolumbar fracture by exceeding the age-determined inherent range of motion and stability of the thoracolumbar spine include flexion, flexion-compression, flexion-rotation, flexion-distraction, extension, lateral shear or slice, and compression ( Fig. 35C-2 ) .


    Jun 11, 2019 | Posted by in ORTHOPEDIC | Comments Off on Thoracolumbar Fractures

    Full access? Get Clinical Tree

    Get Clinical Tree app for offline access