Spinal Fractures and Luxations

Fractures and luxations are classified as traumatic or pathologic. Spinal injury in small animals is most commonly associated with blunt force trauma from vehicular accidents [1-4]. Other traumatic causes include gunshot injuries, encounters with other animals, falling from a height and injuries caused by falling objects [1,3,4]. Pathologic fractures and luxations are often associated with congenital vertebral anomalies, metabolic disorders, and neoplasia. In this chapter we discuss traumatic vertebral fractures and luxations. Secondary spinal cord injury and its treatment are discussed in a separate chapter. Conditions such as atlantoaxial luxation, "wobblers" (caudal cervical spondylomyelopathy), and traumatic disc extrusion are not included. Surgical repair of spinal fractures is discussed with regard to biomechanical properties but techniques are not presented in detail.

Anatomy and Biomechanics of the Normal Spine

The spine is composed of bony structures (the vertebrae) and soft tissues (all associated ligaments, muscles, etc.). Normal spinal motion includes flexion and extension (most important) as well as lateral bending and rotation. Violent forces applied to the spine can lead to fracture or luxation.

Models that divide the spine into bony and soft-tissue compartments have been developed to help predict the stability of various fracture configurations and guide surgeons with respect to treatment. Using a two-compartment theory,5 ventral and dorsal compartments were described. In this model, the ventral compartment consists of the vertebral body, the intervertebral disc (annulus and nucleus pulposus), and the ventral and dorsal longitudinal ligaments. The dorsal compartment contains the dorsal lamina (pedicles and lamina), the articular facets, the joint capsules, the dorsal spinous process, the interspinous and supraspinous ligaments, and the ligamentum flavum. The paraspinal musculature adds stability to both compartments. Lesions that affect both compartments occur most commonly and are generally considered unstable [6]. Traumatic lesions to a single compartment (dorsal or ventral) are rare. When they occur, lesions involving the ventral compartment only are considered more stable than those that involve only the dorsal compartment.

A three-compartment model (three-column) modified from the human literature has also been described [7,8]. In this model, the dorsal compartment remains the same while the structures contained in the ventral compartment are further divided into the following: a middle compartment that contains the dorsal longitudinal ligament, the dorsal half of the disc, and the dorsal half of the vertebral body (basically the floor of the spinal canal), and a ventral compartment that contains the ventral longitudinal ligament, the ventral half of the disc, and the ventral half of the vertebral body [7]. Using the three-compartment model, injuries involving two or more compartments are considered unstable [8].

Biomechanics of Fracture

During trauma, forces are exerted onto the spinal column that can lead to vertebral subluxation, luxation, fracture, or fracture-luxation. Forces associated with specific fracture-luxation patterns include: compression (axial loading), bending (flexion or extension), torsion (rotation), and shear [9]. Although a traumatic event can result in a single force being applied to the spinal column, most fractures result from a combination of forces; flexion and rotational forces are most common [6,8].

Axial loading (or compression) develops when a force is applied along the axis of the spine. Pure axial loading is rare but can occur when a standing dog is struck straight from behind or head-on, creating high compression forces across the intervertebral discs and vertebral bodies. Experimental studies performed in human spinal segments show that increased compression across the vertebral segments leads to increased pressure within the normal intervertebral disc and the vertebral endplates [9]. As the pressure increases, blood is squeezed out of the vertebral endplates, which bulge and eventually crack, allowing nucleus pulposus to be displaced within the vertebral body [9]. The intervertebral discs and vertebral endplates are considered important shock-absorbers of the spine [9].

True compression injuries can occur only in portions of the spine that are in a straight line with the force of trauma at the time of impact. This generally results in a "burst" compression fracture of the vertebral body with or without disruption of the intervertebral disc (Fig. 107-1) [20]. If damaged, the intervertebral disc can extrude dorsally or into the adjacent vertebral body [11]. If the dorsal compartment remains intact, fractures of this configuration are considered relatively stable [10]. However, bony fragments (from the floor of the spinal canal), disc material, and redundant annulus fibrosus can be driven into the spinal canal, causing spinal cord compression. Studies using segments of spine with normal and degenerate intervertebral discs suggest that compression by extrusion of nucleus pulposus occurs only when the disc is abnormal to begin with [9].

Figure 107-1. Compression injuries occur when the affected vertebrae are in a straight line with the force of trauma at the time of impact. This generally results in a "burst" compression fracture of the vertebral body with or without disruption of the intervertebral disc. Pure compression injuries are rare.

Pure hyperflexion and hyperextension injuries are rare. Hyperextension injuries occur when a force is applied directly onto the dorsal aspect of the spine, creating compression forces on the structures of the dorsal compartment and tensile forces on the structures of the ventral compartment [10]. With enough force, the ventral longitudinal ligament and the ventral portion of the annulus fibrosus may rupture, leaving the dorsal ligamentous structures intact. These lesions are stable and often reduce spontaneously, making them difficult to diagnose [10,11].

Pure hyperflexion injuries are rare [10]. Excessive hyperflexion results in a wedge fracture whereby the ventral portion of the vertebral body is crushed, often sparing the ligamentous structures of the spine (Fig. 107-2) [20]. When referring to the three-compartment model, this type of fracture tends to affect only the ventral compartment, leaving the dorsal portion of the disc and vertebral body intact [7]. Sparing of the middle and dorsal compartments prevents subluxation and compression of the spinal cord by bony fragments [7]. These fractures are considered relatively stable [7,10].

Figure 107-2. Excessive hyperflexion of the spine results in a wedge fracture whereby the ventral portion of the vertebral body is crushed, often sparing the ligamentous structures of the spine. Pure hyperflexion forces are rare.

Torsional forces are typically accompanied by flexion of the spine. These forces often cause trauma to bony and ligamentous structures of the ventral and dorsal compartments (two compartment model) and result in unstable fracture-luxations of the spine. When flexion is the primary force applied to the spine and simultaneous rotation occurs, luxation is the typical outcome (Fig. 107-3) [8,11,12]. When rotation is the primary force applied to the spine with concurrent flexion, fracture-luxation generally occurs (Fig. 107-4). Owing to their inherent instability, these lesions are often treated with surgical stabilization [11,13].

Figure 107-3. When flexion is the primary force applied to the spine and simultaneous rotation occurs, luxation is the typical outcome.

Figure 107-4. When rotation is the primary force applied to the spine with concurrent flexion, fracture-luxation generally occurs.

Spinal Cord Trauma

Spinal trauma leads to primary spinal cord injury (concussion, compression, and distraction) caused by the initial impact and by repeated motion at the fracture site. Secondary spinal cord injury involves a series of events that begin soon after the initial injury and include complex vascular, biochemical, and inflammatory processes [8,14]. Because the clinician has no control over the initial spinal cord trauma (primary injury), therapeutic efforts must be aimed at decreasing repeated injury owing to instability and halting or attenuating the secondary processes that begin at the time of injury. The pathophysiology of spinal cord injury and its treatment are discussed in a separate chapter.

Diagnosis

If spinal injury is suspected, patient manipulation should be limited; compliant patients should be strapped to a board to prevent further displacement of an unstable fracture and repeated spinal cord trauma during patient evaluation. A depression or vertebral malalignment along the spine and possible crepitation may be noted on physical examination. A cursory neurologic examination is performed in an attempt to localize the lesion and determine the severity of the spinal cord injury. Sedation of non-compliant patients and administration of analgesics is considered after a rapid assessment of the neurologic status. Concurrent injuries should be taken into consideration when interpreting the neurologic examination to prevent overstating the neurologic dysfunction and prognosis [4,15]. If possible, survey radiographs should be performed with the patient awake or under light sedation in order to maintain paraspinal muscle tone and decrease the risk of further fracture displacement. It is important to remember that radiographs may not necessarily reflect the maximal displacement that occurred at the time of trauma [16,17]. This helps to explain why correlation is poor between radiographic displacement and neurologic deficits [2]. Lateral radiographic views are recommended first and, if available, horizontal-beam ventrodorsal views should also be performed. Survey radiographs of the entire spine are recommended to rule out multiple spinal lesions [1-4,18]. The benefits of performing more thorough radiographic views once the patient is under general anesthesia must be weighed against the potential risks associated with spinal manipulation. Myelography is recommended when no radiographic evidence exists of vertebral displacement and when the neurologic signs do not correlate with the radiographic lesion. In such cases, myelography can help localize the injury by outlining an edematous section of spinal cord or slight vertebral canal subluxation. In addition, myelography can provide information on the degree of spinal cord compression caused by fracture fragments or hematoma. Some authors recommend myelography for all spinal trauma patients [19]. Advanced diagnostic imaging (CT and MRI) is also useful in cases where vertebral displacement is not evident on survey radiographs. Compared with survey radiography, CT provides an accurate assessment of spinal canal compromise [20] as well as superior detail of the fracture lines and bone fragments. This is especially true for fractures that involve the pedicles and articular facets and for displaced bone fragments that may compress the spinal cord. MRI provides good soft-tissue imaging and may help to determine the severity of spinal cord injury, the presence of spinal cord hemorrhage or edema, and whether disc herniation and spinal cord compression are present [21]. As with survey radiography, the risks associated with patient manipulation and positioning under general anesthesia as well as the additional anesthetic time must be weighed against the potential diagnostic value.

All trauma patients, especially those with spinal pain or neurologic dysfunction should be evaluated for possible vertebral fracture or luxation. Patients should be assessed and appropriately treated for other concurrent injuries such as pneumothorax, pulmonary contusions, diaphragmatic hernia, and fractures of the pelvis, ribs, and apendicular skeleton. Previous reports indicate that approximately 50% of patients with spinal trauma have injuries to other systems [1-4], and that approximately 48% of patients with a lumbar fracture or luxation have a fracture elsewhere in the body; most frequently a pelvic fracture [4], Owing to a high rate of concurrent pulmonary injuries, thoracic radiographs are recommended for all patients that are to undergo general anesthesia for fracture fixation [4].

Incidence and Distribution of Spinal Fractures

Spinal fracture-luxations in dogs have been reported to affect the lumbar vertebrae most commonly, followed by sacrococcygeal, thoracic, and cervical vertebrae [2,3,18]. In cats, the sacrococcygeal region is the most common site for vertebral fracture and/or luxation [18]. Stress concentration at or near the junction between mobile and relatively immobile regions of the spine is thought to result in an increased rate of fracture-luxations at the thoracolumbar and lumbosacral junctions 18]; however, studies have also described a relatively even distribution of fractures along the lumbar spine.4,22

The vertebral body has been identified as the most commonly fractured portion of the vertebrae [2,18]. Disc herniation associated with fractures or luxations is fairly uncommon and was reported in fewer than 12% of cases in one study [18]. The presence of more than one spinal lesion has been reported, but this is rare and tends to involve two adjacent vertebrae [18]. These lesions most commonly involve the lumbosacral and sacrococcygeal junctions [18]. Despite their rare occurrence, radiographic evaluation of the entire spine is recommended to rule out multiple lesions [1-4,18].

Cervical Fractures

Cervical fractures and luxations are associated with various degrees of neurologic dysfunction but do not typically lead to loss of pain perception caudal to the lesion. The main reasons for this may be that: 1) the larger cervical spinal-canal to spinal-cord ratio allows for more displacement without severely damaging the spinal cord, and 2) cervical trauma severe enough to cause loss of pain perception generally causes respiratory failure and death at the time of injury [12,23 ].Although some have reported that most dogs with cervical fractures present with mild neurologic deficits and often only neck pain [23], a study reviewing 56 cases of cervical fracture-luxation revealed that 57% of cases presented with non-ambulatory tetraparesis [1].

The axis (C2) is most frequently fractured (50-78% of cases) followed by the atlas (C1) (approximately 25% of cases) [1,2,23]. The higher incidence of C2 fractures in dogs is thought to relate to the anatomy of the proximal cervical region. It has been hypothesized that the axis acts as a point of leverage between the cranial and caudal cervical regions. This is thought to be due to the close relationship of the vertebral body of the axis with the atlas and the intimate soft-tissue attachment of the dorsal spinous process and caudal articular facets of the axis with the caudal cervical vertebrae [12,23]. One study reports that fractures of C3-7 are more common after dog fights or unknown trauma than after being hit by a car and that fractures involving the caudal cervical vertebrae typically affect more than one vertebra [1].

High perioperative mortality (36%) has been reported with surgically treated cervical fracture-luxations [1]. However, most patients that survive through surgery and the immediate postoperative period achieve functional recovery [1]. In addition, almost 90% of patients treated conservatively obtain a functional outcome, suggesting that many cervical fractures are amenable to conservative therapy [1]. Negative predictors for recovery include a non-ambulatory status at presentation and being referred more than 5 days after the initial trauma [1].

Thoracic and Lumbar Fractures

The lumbar and thoracic regions of the spine are common sites for vertebral fracture and/or luxation in dogs and cats [2,3,18]. Although fractures reportedly occur more commonly at the thoracolumbar and lumbosacral junctions [18], a study by Turner and others revealed that fractures of L2, L4, and L7 were more common than fractures of the other lumbar vertebrae; no statistical difference was found when all lumbar vertebral segments were compared [4]. Fractures between T1 and T9 are less common, typically minimally displaced, and relatively stable. The inherent stability of this section of the spine is thought to be due to the additional support provided by the ribs and their attachments. Neurologic dysfunction varying from alterations in conscious proprioception to loss of pain perception has been reported in as many as 85% of patients with lumbar fractures [4]. Based on previous reports, 45% of patients with lesions between T1 and S1 and as many as 62% of patients with lumbar injuries were euthanized without treatment [3,4]. The decision to euthanize was often based on the severity of the neurologic dysfunction or on surgical confirmation of severe spinal cord injury [3,4]. The high rate of neurologic dysfunction associated with thoracolumbar injuries compared with cervical injuries is likely related to the relatively small spinal-canal to spinal-cord ratio of the thoracolumbar spine [15]. In contrast, fracture-luxations of L6 and L7 only affect nerve roots (cauda equina); these can tolerate a greater amount of displacement without severe neurologic dysfunction [24]. Unlike fractures of the lower thoracic and lumbar regions, which often require surgical fixation, fractures of L6, L7, and the lumbosacral region often heal satisfactorily without surgical intervention [15].

Sacrococcygeal Fractures

Sacrococcygeal fractures are the most common spinal fracture diagnosed in cats [18,25], and they occur relatively frequently in dogs [18,26]. Clinical signs associated with these fractures range from sacral pain to loss of sensation and motor function to the tail and perineal area, along with urinary and fecal dysfunction. Although neurologic deficits to one or both hind limbs are possible with traction lesions (tethering) of the cauda equina and spinal cord, these lesions are rare and difficult to differentiate from those associated with concurrent pelvic fractures [26,27]. In dogs, sacral fractures that occur medial to the foramina (axial) are more frequently associated with urinary or fecal incontinence, loss of perineal sensation, and tail analgesia than are sacral fractures that occur lateral to the foramina (abaxial) [26]. In addition, dogs with axial fractures have more severe neurologic dysfunction at presentation and discharge than do dogs with abaxial fractures [26]. Additional orthopedic injuries are reported in 76 to 88% of cats and 74 to 97% of dogs, which emphasizes the need for a complete assessment in order to provide an accurate prognosis [25-27]. In cats, sacral fractures are often associated with sacroiliac luxation, whereas in dogs, they are more commonly associated with ilial fracture [27]. More than 70% of cats in one study had temporary or permanent urinary incontinence, but the prognosis for recovery of normal urinary function was good when cats retained anal tone and perineal sensation at presentation [25]. Return of neurologic function to the hind limbs, bladder, and rectum reportedly occurs within 1 month; cats that have not recovered urinary function 1 month after trauma are unlikely to recover [25]. In a canine study, patients with abaxial fractures had a better prognosis for recovery than those with axial fractures [26]. Based on the results of this study, little neurologic improvement should be expected during hospitalization [26]. Surgical treatment of sacrococcygeal fractures has been associated with worsening of the neurologic status in approximately 40% of dogs; this is likely owing to iatrogenic injury [26].

Principles of Fracture Fixation

Whether to treat spinal fractures conservatively or surgically is a topic of great controversy. Surgical intervention aims to reduce and realign vertebral bone fragments, which in most cases will effectively decompress the spinal cord; and to provide rigid fixation of the affected vertebral segments to prevent continued instability. In some cases, laminectomy is necessary to assess the spinal cord for visible damage and to relieve compression caused by hematoma, disc material, or bony fragments. Because decompressive procedures have the disadvantage of potentially destabilizing the spine, pediculectomy and hemilaminectomy are preferred over dorsal laminectomy because they produce less instability [2,28]. Although the general consensus is that fractures that involve both the dorsal and ventral compartments (two-compartment model) are unstable and should be stabilized surgically, some studies have demonstrated up to 94.4% functional recovery (walking) with non-surgical management [1,2,22]. Although patients with a range of neurologic dysfunction were included in published retrospective studies, reported recovery rates should be interpreted with caution because patients with severe neurologic dysfunction are more likely to have undergone surgical stabilization, whereas patients with less severe neurologic dysfunction are more likely to have been treated conservatively. Despite this, patients retaining even minimal neurologic function should be given the benefit of conservative management if surgery is not an option and the owners are willing to perform supportive care and rehabilitation [3,22].

Conservative management and external splinting have demonstrated good success in the treatment of some cervical and thoracolumbar injuries [3,22,29]. Because it provides relatively limited immobilization of the fracture site, this treatment option has typically been recommended for patients with relatively stable fractures causing mild neurologic dysfunction or when financial constraints limit treatment options [5]. Experimental bending of a back splint designed for the treatment of lower thoracic and lumbar fractures revealed it could withstand bending moments in excess of those shown to cause failure of commonly used internal fixation techniques and those experienced by large paralyzed veterinary patients undergoing nursing care [29]. This splinting technique is considered suitable for patients that retain some degree of voluntary motor function and have lesions that affect mostly the dorsal compartment; lesions with loss of ventral buttress are not considered ideal for this technique [29]. External splints should extend well beyond the fracture site on either side of the affected vertebral segments to prevent a fulcrum effect at the level of the fracture. Patients should be monitored daily for slippage of the splint and other related complications such as urine scalding, skin abrasion, ulceration or abscessation, and overheating [22,29]. Although conservative management requires significantly more supportive care and generally leads to a longer recovery period, the cost of treatment is often lower and the duration of hospitalization shorter than for surgical stabilization [3].

Surgical management should be considered for exploratory purposes in patients that have lost pain sensation, and in order to provide stable fixation across the fracture site in all patients with severe neurologic deficits (non-ambulatory patients) [25,30]. Surgical stabilization is also recommended for patients with unstable fractures (determined radiographically or by palpation of a "click" during ambulation or patient movement), those that fail to improve or continue to deteriorate neurologically despite conservative therapy, and patients with severe pain beyond the first 48 to 72 hours after injury [25,30].

Various surgical techniques have been described for fracture stabilization in all regions of the spine. Vertebral body cross-pinning was described for stabilizing fractures of the articular facets, epiphyseal body fractures, and spinal luxations. This technique is considered inadequate for fixation of severe compression or multifragmented fractures [31] and is no longer recommended. Spinal stapling involves applying parallel stainless steel pins through and along the dorsal spinous processes on either side of the fracture, with additional wire stabilization through the dorsal spinous processes spanned by the repair [32]. This technique is only applicable for patients weighing less than 10 kg and requires intact dorsal spinous processes [5,32]. Similarly, modified segmental spinal instrumentation involves placing Steinman pins around and along the dorsal spinous processes of several vertebrae and wiring these pins to the dorsal spinous processes and the cranial articular facets of the vertebrae spanned by the repair [33]. This technique is reportedly applicable for all sizes of dogs but should be reserved for fractures that do not involve the ventral compartment. Modified segmental spinal instrumentation and spinal stapling frequently fail by tearing through the spinous processes and are rarely used. Application of a metal bone plate to the dorsolateral aspect of two or three consecutive vertebral bodies is referred to as vertebral body plating. This technique is limited to the caudal thoracic and cranial lumbar vertebrae because application may require dislocation or removal of any associated rib head and transection of nerve roots (rhizotomy) exiting below the bone plate, precluding its use in the lower lumbar spine [34].

Dorsal spinous-process plating using plastic (Lubra plate, Lubra Co., Fort Collins CO 80521, USA) or metal (Auburn spinal plate, Richard Manufacturing Co., Memphis, TN 38101, USA) spinal plates has been described as an option to stabilize dorsal compartment instability in regions with prominent dorsal spinous processes [30,35,36]. This technique has been used extensively but appears to be of limited use on its own. Dorsal spinal plating is most commonly applied in combination with other methods such as transilial pinning and vertebral-body plating to stabilize fractures of the lumbar vertebrae that do not involve the dorsal spinous process or lamina [30,35,37]. Reported complications include fracture of the dorsal spinous processes, ischemia of the dorsal spinous processes as a result of over-tightening, and plate slippage owing to insufficient tightening of the bolts [30,38]. The use of screws or pins plus polymethylmethacrylate (PMMA) has been described for stabilization of spinal fractures in all locations, including those that are less amenable to ventral compartment fixation such as the thoracic and lower lumbar region [39-42]. This procedure requires minimal instrumentation, however, an excellent knowledge of vertebral anatomy is essential to ensure accurate implant positioning. This technique is compatible with decompressive techniques as long as the surgeon carefully applies and irrigates the PMMA during polymerization. Pins and PMMA, unlike several other stabilization techniques, can be applied even when perfect alignment is not achieved [2]. Stabilization using pins and PMMA is currently the most popular technique used to stabilize vertebral fractures. External skeletal fixation has been reported for stabilization of lumbar fractures [43,44]. Advantages include that implants are inserted away from the fracture site and that removal of all implants is possible once healing has occurred [43,44]. Again, excellent knowledge of vertebral anatomy is essential to ensure accurate placement of pins into the vertebral bodies; this can be facilitated by fluoroscopic guidance [45].

Experimental comparison of the biomechanical characteristics of various surgical techniques used to stabilize spinal fractures has been performed. A drawback of these studies is that they typically evaluate a small number of fixation techniques and cannot be compared with each other because differing methods are used for each evaluation.

It is well accepted that techniques providing stability to both the ventral and dorsal compartments are superior to those that stabilize only a single compartment. When the stability and strength of five internal fixation methods used for lumbar fracture stabilization were compared, vertebral body plating was found to be stronger than all other methods including dorsal spinal plating, pins and methylmethacrylate, and crossed pins [46]. The combination of dorsal spinous-process plating and dorsolateral vertebral-body plating provides the most rigid and strong repair [46]. Similarly, a more recent study comparing five methods of fixation for thoracolumbar instability reported that the application of vertebral body plates or the combination of pins and PMMA was strongest [47]. A biomechanical study comparing five configurations of pins or bone screws with PMMA for internal stabilization of lumbar instability revealed that 8 pin fixations were more rigid and stronger than 4 pin fixations and that the orientation of the pins impacted on rigidity depending on the number of pins used [48]. Furthermore, this study revealed that pins were more rigid and less likely to fail than screws [48]. Finally, a study showed that external skeletal fixation has mechanical properties comparable to internal fixation techniques that use a combination of pins and PMMA [44].

Prognosis and Recovery

Similar to other spinal disorders, the loss of pain perception caudal to the lesion is the single most accurate prognostic indicator when evaluating patients with spinal fracture and/or luxations [5,16,49]. Patients that have no deep pain have less than 50% chance of recovery, and the rate of recovery decreases to less than 5% if pain sensation has been lost for more than 48 hours [50]. Despite having a positive impact on the prognosis, the presence of pain perception does not guarantee complete neurologic recovery.

Based on previous reports, it appears that the degree of radiographic vertebral displacement does not correlate well with the neurologic status [2,16]. Although survey radiographs are of limited use when trying to determine prognosis in patients with little vertebral displacement, correlations have been made between the degree of radiographic vertebral malalignment and the loss of pain perception caudal to the injury [16,22]. Patients with 80% vertebral displacement in the thoracic spine and 60% displacement between L1 and L5 do not typically maintain pain perception caudal to the lesion [16]. Caudal to L5, survey radiographs are less helpful in determining outcome because the nerve roots at this level can tolerate significant displacement without severe neurologic dysfunction [16]. In general, vertebral displacement of greater than 80% caudal to L5 is associated with LMN signs [16]. Unlike the cervical and thoracolumbar regions, full recovery is possible in patients with severe vertebral displacement if the lesion is caudal to L5 [16].

Patient recovery is difficult to predict, although patients that retain pain perception and purposeful movement are likely to recover functional neurologic status. Based on previous reports, the hospitalization time for surgically treated cases tends to be longer than for cases treated conservatively [3]. However, the overall rate of return to a functional neurologic status is similar for both treatment regimens [2,3].