Anatomy and Biomechanics

Fractures of the thoracic spine, defined as injuries of the first through tenth thoracic vertebrae, are quite commonly seen in patients, and may be divided into low-energy (including pathological fractures)1,2 and high-energy subtypes.3 This bimodal energy distribution seen in trauma may be due to the kyphotic attitude of the thoracic spine4 and the stabilizing effect of the rib cage and sternum—protection not afforded to other segments of the spine.5–7 Andriacchi et al8 calculated that the compression tolerance of the thoracic spine is increased four times by the presence of an intact rib cage and sternum. The buttress of the rib cage, along with the coronally oriented thoracic facet joints, likely combine to limit transverse plane rotations in the thoracic spine as compared with the cervical and lumbar segments.5 This means that substantial energy transfer is required to cause many traumatic thoracic fractures. Conversely, loss of the sternal stabilizer can be significant. Lund et al9 found that even relatively trivial wedge compression fractures may signal an unstable injury pattern in the presence of a sternal fracture. As with cervical spine injuries, there is a nontrivial rate (17%) of nonadjacent spinal injury when thoracic spine injury is diagnosed,10 so the importance of thorough spinal evaluation must be emphasized. For these injuries, clinical examination is likely insufficient, and should be supplemented with advanced imaging such as a computed tomography (CT) scan.11

In contrast to the thoracic segment, the thoracolumbar spine (T10 to L1) is not nearly as buttressed, as the ribs are not articulating with the sternum, and the facet joints begin to transition in morphology, becoming more curved and obtaining more of a sagittal-plane orientation. This may contribute to the observed higher prevalence of spinal injuries in this transitional zone.

The vital organs ventral to the thoracic spine (lungs, aorta, diaphragm, duodenum, spleen) that can be injured concurrently with the spine often complicate and inform treatment options for these injuries.12 We explore some of these injuries in more detail later in the chapter.

Injury Patterns and Classification

Isolated spinous process fractures (e.g., “clay-shoveler′s fracture”) generally represent avulsion injuries often seen just cranially to osteoporotic vertebral compression fractures.13 They can generally be treated conservatively, requiring no special restrictions other than those imposed by pain until healing.14 This has been shown to be the case for isolated thoracic transverse process fractures as well.15 Prolonged late pain from a mobile spinous process fracture may be treated with isolated resection.

Thoracic and thoracolumbar classification systems have been proposed since the inception of roentgenograms, and have generally had three purposes: to enable physicians to accurately communicate what structures have failed, to assist practitioners in guiding treatment, and to help determine the prognoses for these injuries. To date, all the classification systems have fallen short in one aspect or another. However, a brief synopsis of their historical evolution can be useful. Watson-Jones16 asserted that “perfect recovery is possible only if perfect reduction is insisted upon.” His method of postural reduction focused on counteracting the predominant kyphotic deformity of these injuries (Fig. 11.1).

In 1949, Nicoll17 classified the fractures that occurred in the coal miners he cared for. He felt that stability was predicated on intact interspinous ligaments, and he stated, interestingly, that functional recovery was not dependent on achieving anatomic reduction of the fracture. In the 1960s, Holdsworth18,19 similarly proposed that the posterior ligaments were the key to stability. Whitesides20 described a two-column theory of the thoracolumbar spine in which the anterior structures (vertebral bodies and disks) are subjected to primarily compressive stresses, whereas the posterior column (laminae, transverse and spinous processes, joint capsules) is subjected to primarily tensile stresses (Fig. 11.2). He correctly noted that the center of mass for a thoracic vertebral body lies ventral to the spine, and compared the spine to a construction crane.

Denis21 studied CT scans of thoracolumbar fracture and proposed a biomechanical three-column model in which the posterior disk and posterior longitudinal ligament compose the middle column. His description of four common injury patterns—compression (48%), burst fracture (14%), fracture-dislocations (16%), and seat-belt injuries (5%)—is still commonly used. McCormack and Gaines′s group22 developed a classification that quantified the remaining structural integrity of a burst fracture to predict when short–segment pedicle screw fixation would be sufficient treatment.

Magerl et al23 described three basic types of thoracolumbar injury patterns (commonly known as the AO classification) based on their series of 1,445 fractures and their perception of the primary mode of failure, as follows: type A, compression; type B, distraction injury to both columns; and type C, rotational injury to both columns. Vaccaro et al24,25 built on the AO classification using a method that assesses (1) the mechanism of injury, (2) the neurology, and (3) the function of the posterior ligamentous complex (PLC). Points are assigned based on injury morphology, integrity of the PLC, and neurologic status (Table 11.1). Point totals < 4 are recommended for nonoperative care, those with a point total > 4 are recommended for surgical treatment, and those with a point total of 4 are indeterminate.

To this end, Crosby et al26 demonstrated that sagittal short T1 inversion recovery (STIR) magnetic resonance imaging (MRI) is relatively sensitive and specific in determining injury to the PLC, with greater than 90% agreement among surgeons of different levels of experience. We have found this to be a useful system when discussing injuries with our medical students and residents. It has recently been shown to correlate with the AO classification and to be predictive of current treatment strategies.27

Neurologic Injury

As the spinal cord courses through the thoracic spine, surgeons who care for these injuries must also be conversant with concomitant spinal cord injuries (SCIs). Meyer28 noted that 63% of the thoracic spinal injuries seen in his practice had associated complete neurologic injuries. However, the incidence of concomitant SCI may be decreasing with improvements in motor vehicle safety.29 During resuscitation, it must be remembered that 5% of thoracic spinal cord–injured patients may manifest neurogenic shock, and will require support in the form of fluids and pressors.30 Certainly, the severity of initial SCI will portend the recovery prognosis.31–35 Thus, for this reason, and others, it is strongly recommended that practitioners quantify the patient′s neurologic status on admission and daily thereafter using the American Spinal Injury Association (ASIA) Impairment Scale (Fig. 11.3).36 One can also categorize overall spinal cord function according to the modified Frankel score,36 and we find this very helpful to quickly convey neurologic status in our thoracic–injured patients (Fig. 11.4).

Harrop et al37 recently reported the 10-year experience of patients treated at their trauma center with thoracic-level SCIs, and suggested that injury level also impacts prognosis. They found that injuries at levels T4 to T9 had the least potential for recovery, whereas those at levels T10 to T12 demonstrated a much higher rate of recovery. They also showed that those with a complete neurological injury (ASIA category A) had only a 4% rate of improvement, whereas 70% of those with incomplete injuries (ASIA categories B to D) demonstrated some improvement. It should also be noted that those with high-thoracic SCI are at greater risk of pneumonia and death.38 The optimal timing of surgical decompression and stabilization continues to be a topic of much discussion, but remains without a strong clinical evidence base39,40 as regards its effect on ultimate neurologic recovery. However, it has been shown that early fixation of thoracic fractures reduced respiratory morbidity as well as the number of days on ventilator and the number of days in the intensive care unit (ICU).41,42 Kerwin et al43 have proposed that thoracic spine fractures be stabilized operatively (when necessary) within 72 hours of injury.

Source: Adapted from Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S. A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 1994;3:184–201; and Vaccaro AR, Lehman RA Jr, Hurlbert RJ, et al. A new classification of thoracolumbar injuries: the importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine 2005;30: 2325–2333.

Current Treatment Options

The goals of treatment for thoracic fractures are (1) prevent additional damage to the neural elements through careful resuscitation of the patient and surgical decompression of the neural elements; and (2) provide mechanical stability to the axial skeleton when necessary, to further prevent injury and to enable rapid, safe mobilization of the patient. Mechanical stability is often achieved through a period of enforced recumbency, casting/bracing (if the body habitus allows, i.e., nonobese patients), or the use of internal fixation (anterior, posterior, or both). Surgeons must also assess the need for anterior column reconstruction, and then craft a fixation strategy that requires fusion of the minimum number of spinal segments. One area of continued treatment controversy is the neurologically intact thoracolumbar burst fracture. Although there can be no doubt that a subset of these injuries are unstable, rigorous studies continue to show only small improvements in outcomes with operative fixation at the cost of increased complication rates.44

Operative Treatment of Thoracic Fractures: Overview

Surgical treatment for thoracic fractures is typically reserved for those with associated neurologic injury and compression of the neural elements, or instability (the concern that the injured segment has lost enough mechanical integrity that the spinal cord is at risk of new injury under anticipated physiological loading). Surgical treatments thus must accomplish decompressions and/or stabilization. Some fractures may require surgery in a subacute setting secondary to delays in diagnosis, persistent pain, or deformity after attempted nonoperative care.

The specific approach and extent of surgery will vary based on the amount of perceived instability, the need for decompression, as well as the skills and experience of the surgeon. Injuries with significant thoracic cord compression from ventral bone shards will typically require some version of ventral decompression, as the thoracic spinal cord cannot be safely retracted to facilitate a direct dorsal approach. Laminectomy alone should generally be avoided in the thoracic spine.30,31,48 Posterior decompression alone cannot be expected to reliably decompress the spinal cord if the lesion is ventral; the cord in the kyphotic thoracic segment cannot be expected to “float back” away from ventral compression as it might in the cervical spine. For this reason surgeons have described decompression and reconstruction through thoracotomy49 and posterolateral retropleural50,51 approaches. These approaches to directly decompressing the thoracic spine are illustrated in Fig. 11.5. Lubelski et al52 performed a meta-analysis of the literature describing complications from various thoracic spine approaches, and demonstrated a 39% complication rate when thoracotomy was employed, whereas costotransversectomy incurred only a 15% complication rate. The underlying reasons for these differences were not entirely clear.

Mechanical stability and sagittal plane alignment issues will also inform the choice of surgical plan. Edwards and Levine53 recognized that Harrington distraction rods were suboptimal in this regard for thoracic fractures, and proposed the addition of a rod sleeve to apply a three-point bending extension moment to the injured segment. Later, posterior pedicle screw fixation constructs demonstrated superiority in conferring stability and promoting fusion in these patients.54,55 Pedicle screw attachment has become the dominant method of posterior fixation, but questions of when this is mechanically sufficient remain. Fractures with minimal comminution of the anterior column will typically be more stable due to inherent load-sharing and can often be treated with a short-segment posterior construct.22 Fractures with a high degree of anterior comminution typically need an anterior column reconstruction or a longer posterior construct to compensate for the loss of load sharing anteriorly. Fracture-dislocations typically require the most robust constructs to regain stability.

Vertebroplasty and Kyphoplasty

The indications for these procedures continue to evolve and are controversial, but it is currently an accepted treatment for a symptomatic low-energy compression fracture that has failed 4 to 6 weeks of conservative care (i.e., analgesics, activity modification, bracing). Vertebroplasty or kyphoplasty might also be indicated in the acute setting where pain due to a compression fracture is intractable and the sole barrier to mobilizing the patient. Other considerations would be in the setting of an otherwise stable pathological fracture such as one due to multiple myeloma or metastatic disease. For vertebroplasty, a cannula is placed into the body via a posterior transpedicular approach, and cement is injected into the body under fluoroscopic guidance to stabilize the fracture fragments. Kyphoplasty requires the same approach, but prior to injecting cement a balloon is inflated in an attempt to reduce the compressed end plate and to restore normal sagittal profile. This same void in the vertebral body created by the balloon is then filled with cement (Fig. 11.6).

Studies directly comparing vertebroplasty to kyphoplasty have failed to show any significant difference in post-procedural pain relief.56,57 This finding, along with the substantially higher cost of kyphoplasty, has led many surgeons to prefer the vertebroplasty procedure. Proponents of kyphoplasty cite the potential for restoring the sagittal profile as a major benefit and rationale for its use, although this has not been substantiated. Two randomized, placebo-controlled prospective studies published concurrently in 2009 called into question the treatment effect of vertebroplasty, finding it indistinguishable from placebo (sham procedure).58,59 This led to a serious reexamination of the value of vertebral augmentation procedures. However, there were multiple problems with the design and execution of these studies. Klazen et al60 subsequently published results that demonstrated the efficacy of vertebroplasty in acute compression fractures with persistent pain, and this relief was maintained at 1 year. More recently, Farrokhi et al61 published similar effectiveness for vertebroplasty, and Edidin et al62 demonstrated increased life expectancy in the Medicare population when treated for compression fracture with vertebral augmentation.

Related posts:

Treatment of Fractures in Geriatric Patients Care of the Soft Tissue Envelope Musculoskeletal Infection Associated with Skeletal Trauma Distal Humeral Fractures Femoral Neck Fractures Pelvic Ring Injuries

Stay updated, free articles. Join our Telegram channel

Join