Traumatic thoracic spinal fracture dislocation with minimal or no cord injury

J Neurosurg (Spine 3) 96:333 337, 2002 Traumatic thoracic spinal fracture dislocation with minimal or no cord injury Report of four cases and review of the literature SCOTT SHAPIRO M.D., TODD ABEL, M.D., AND RICHARD B. RODGERS, M.D. Section of Neurosurgery, Indiana University Medical Center, Indianapolis, Indiana Object. Thoracic fracture dislocations reportedly lead to complete paraplegia in 80% of cases. It is rare for these dislocations not to cause neurological deficits, as evidenced by the mere 11 well-documented neurologically intact cases in the English-language literature. Methods. The authors report four cases of thoracic fracture dislocation that remained neurologically intact and discuss 11 other previously published well-documented cases. Overall there were 10 men and five women with a mean age of 34 years (range 17 66 years). Mechanisms of injury included car crash in six, motorcycle crash in seven, plane crash in one, and fall from a horse in one. On admission, neurological deficits were absent in 11 patients, intercostal neuralgia was present in two, and mild lower-extremity weakness/numbness was demonstrated in two. All suffered significant thoracic pain, with 14 patients having sustained rib fractures and eight a hemothorax. The levels of dislocation were T3 4 in two, T5 6 in four, T6 7 in four, T7 8 in two, T8 9 in one, and T9 10 in two. All suffered some degree of lateral translation (mean 12 mm, range 3 27 mm). There were six cases of well-documented anterior subluxation in addition to translation (mean 12 mm, range 4 23 mm), and all involved some degree of fracture imploding of one vertebral body (VB) into an adjacent VB. There were six cases of burst fracture with translation (mean kyphotic angle 38, range 28 50 ). Bilateral pedicle shear fractures were present in all 15 cases at the site of subluxation, thus separating the anterior from the posterior elements and preserving the spinal canal. Only two of the 15 patients suffered complete spondylolisthesis. Five patients underwent successful nonoperative management with prolonged bed rest; at follow-up examination, neurological status remained normal in all five, lesions were radiographically unchanged in three, and there was less subluxation but not anatomical alignment in two. Ten patients underwent successful internal fixation via anterior approaches (two cases), posterior approaches (five cases), and combined approaches (three cases). Neurological status either improved to normal or remained normal except in one case with persistent intercostal neuralgia. Surgery resulted in no change in alignment in three, improved but not anatomical alignment in 11, and normal alignment in one patient. All patients ambulated unassisted by 6 months. Conclusions. In cases in which bilateral pedicle fractures occur at the site of significant thoracic subluxation and/or translation, preservation of the spinal canal and spinal cord neurological function can rarely occur when both the lamina and spinal cord do not dislocate along with the anterior VBs. In these instances, perfect anatomical reduction may require forces that unnecessarily put neurological function at risk and the results appear to justify internal fixation with some or no reduction of deformity. KEY WORDS spinal fracture spinal stabilization thoracic spine T HE thoracic spine is rigidly stabilized by strong anterior and posterior longitudinal ligaments, the ribs with their costotransverse ligaments and articulation with both the spine and manubrium, a thick ligamentum flavum, and a sagittal orientation of the facet joints that resists axial rotation and horizontal translation. 1,2,5 Thus, it takes a significant force of injury to cause a fracture dislocation in the thoracic spine. A force vector that Abbreviations used in this paper: CT = computerized tomography; MR = magnetic resonance; MVA = motor vehicle accident; TSRH = Texas Scottish Rite Hospital; VB = vertebral body. can cause such a fracture, when combined with the facts that the thoracic spinal canal is narrow and the blood supply to the thoracic spinal cord is fragile, reportedly leads to an 80% incidence of complete paraplegia. 2 Cases of thoracic fracture dislocation in which normal or near-normal neurological status is demonstrated at presentation are rare. Even rarer are cases of complete traumatic thoracic spondylolisthesis in which neurological function is normal. The occurrence of this problem, however, may be more common in the modern era of better driver-restraint systems as well as our improved ability to extricate victims from accident sites. The management of spinal instability and or cord compression takes on more significance 333

S. Shapiro, T. Abel, and R. B. Rodgers in this subgroup. Since 1987, we have treated four cases of thoracic fracture dislocation in patients with normal or near-normal postinjury neurological status; this makes up 3% of all thoracic fracture dislocations that we have managed. We report our experience and analyze similar cases in the English-language literature. Case Reports Case 1 This 23-year-old man involved in an MVA presented with thoracic pain and normal neurological status on examination. Plain x-ray films of the thoracic spine and a CT scan (Fig. 1) demonstrated spondylolisthesis at T5 6 with multiple pedicle fractures, laminar fractures, transverse process fractures, and spinal canal preservation. An MR image (Fig. 2) revealed spinal cord preservation at the level of spondylolisthesis. Because of the extensive nature of bilateral transverse process and pedicle fractures, and laminar fractures at many levels above and below the level of spondylolisthesis, which precluded secure posterior hook or screw placement, as well as the extensive separation of the thoracic anterior elements from the posterior elements, it was elected initially to prescribe 3 months of complete bed rest to allow for bone healing. At 3 months after presentation, a three-dimensional CT reconstruction demonstrated enough callous formation and bone healing to allow placement of secure spinal instrumentation (Fig. 3A and B) but not enough, in our opinion, to ensure stability. The patient underwent surgery via a left transthoracic approach and placement of TSRH VB screws in T4 8, connected to a 0.25-in rod; and a morselized rib graft was also placed. The patient also underwent placement of a posterior bilateral TSRH hook-and-rod construct, and a morselized rib graft from T-4 to T-8 (Fig. 4). No surgical reduction was attempted. The patient remained neurologically normal and was working full time at 5-year followup examination. Case 2 This 66-year-old woman involved in an MVA presented with normal strength and mild hypalgesia below T-4. Thoracic radiography revealed T3 4 VB fracture subluxation with translation to the left (Fig. 5A). Computerized tomography and MR imaging (Fig. 5B) demonstrated extensive fracturing of all three columns of both T-3 and T-4 as well as severe spinal canal narrowing and cord compression. The patient underwent a left thoracotomy with T-3 and T-4 corpectomies; banked fibula graft assisted locking plate internal fixation was performed; this was followed by a posterior T3 4 laminectomy and T1 6 TSRH bilateral hook-and rod claw construct with autologous iliac crest fusion. Her sensory status improved to normal. At 5-year follow-up examination, her sensory status remained normal, and she was working. Case 3 This 34-year-old man was ejected from his vehicle during an MVA. He complained of back pain and his right leg sustained 4/5 diffuse motor weakness as well as hypalgesia. A thoracic CT scan obtained 6 days postinjury FIG. 1. Case 1. Axial CT scans. A: Scan obtained at upper T5 6, documenting pedicle disruption from the T-6 VB and right T-5 with preservation of the spinal canal. B: Mid-T5 6 scan documenting shearing of pedicles from left T-6 and right T-5 with side-by-side VBs and canal preservation. C: Pedicle and laminar/pedicle fractures at T-4, separating the preserved spinal canal and posterior elements from the VB. D: Shear fracture through the T-7 VB, separating the preserved spinal canal and posterior elements from the VB subluxation. demonstrated a fracture translation of T-5, 2 mm to left of T-6, telescoping of a bone fragment from the facet beneath the lamina that compressed the spinal cord, and an isodense epidural hematoma (Fig. 6). The patient underwent a decompressive laminectomy, hook-and-claw TSRH internal fixation, and lateral mass fusion with no reduction. His status recovered to normal, and he was discharged 8 days postoperatively. His motor strength has remained normal and he has been working for 6 years. Case 4 This 54-year-old woman was involved in a motorcycle accident. Thoracic spine radiography revealed a T4 5 fracture dislocation. Her neurological status was normal. A CT scan demonstrated that T-4 was translated laterally by 9 mm and and anteriorly by 6 mm into the T-5 VB with a resultant 30 of kyphosis. The T-5 VB was extensively fractured as well as the pedicles of T-4 and T-5. She underwent a left thoracotomy, T-5 corpectomy, cadaveric fibu- 334

Thoracic spinal fracture dislocation FIG. 2. Case 1. Axial MR image demonstrating spinal cord rotation (arrow) and compression of the anterior subarachnoid space but normal spinal cord in the middle of the juxtaposed VBs of T-5 and T-6. la assisted reconstruction, and Z-plate internal fixation. At 3 years her status has remained normal. Discussion Including the aforementioned four cases, there are 15 well-documented cases of thoracic fracture dislocation in which normal or near-normal neurological function was demonstrated. 4,6 10,12,14,16 18 Although there are additional reports, few if any details were provided. Thus, it is clear that noncord-injured thoracic fracture dislocation is very uncommon, comprising 3% of our experience since 1987. A summary of the findings of all 15 well documented FIG. 3. Case 1. A: Anterior CT reconstruction obtained after 3 months of bed rest, documenting spondylolisthesis and callous formation. B: Lateral CT scan reconstruction documenting spondylolisthesis and callous formation. FIG. 4. Case 1. Anteroposterior radiograph obtained after anterior posterior TSRH-assisted T4 8 reconstruction. cases is shown in Table 1. All 15 fracture dislocations required a great deal of force to occur. In a thorough study of the biomechanics of spinal injuries, Roaf 13 found the spine to be highly vulnerable to rotational and shear forces. He reported that hyperflexion or -extension alone is unlikely to create a fracture dislocation but that in association with a rotational force, failure of all three columns becomes far more likely. Depending on the additional compressive impact, fractures of the VBs may occur. The latter have been described as burst dislocations by Hanley and Eskay. 7 In most cases, it is the inferiorly involved vertebra that bursts, such as in our Case 1. Bohler 1 subdivided fracture dislocations into those with translational or rotational dislocation without fracture of the neural arch (lamina and pedicle), which was associated with a high incidence of paraplegia, and injuries with fracture of the posterior elements called the saving fracture of the neural arch, which led to canal and neural preservation. As previously reported, most thoracic fracture dislocations with neural sparing occurred between T-6 and T-9. 10 Miyasaka, et al., 12 explained this occurrence by the fact that in this area the spinous processes extend farther inferiorly than in any other part of the spine. Consequently, they stated, strong shear forces would be concentrated in the middle column, leading to fractures of either pedicle and preservation of the spinal canal. The uniform concept present in all 15 cases is one of pedicle and/or laminar fracturing that allows for separation of the anterior and middle columns from the posterior column; thus, the spinal cord is not catastrophically stretched or compressed by the VB subluxation. The MR images shown in this 335

S. Shapiro, T. Abel, and R. B. Rodgers FIG. 5. Case 2. A: Anteroposterior radiograph documenting T3 4 fracture dislocation with translation. The upper arrow indicates the T-3 spinous process translated to the left of the T-4 spinous process. The lower arrow indicates the spinous process and mid-vb of T-6. B: Sagittal MR images documenting T3 4 VB destruction with associated spinal cord compression. report, however, demonstrated interesting deformity of the cord with normal neurological function. There have been numerous reports of the high degree of instability in cases with this type of injury. 2,3,5,11 Interestingly, in five of the well-documented cases the patients were successfully treated with bed rest, halo femoral traction, and/or placement of casts (Table 2). 6,12,16,17 The greater the degree of translation, subluxation, and kyphosis, the more likely that conservative therapy will fail. 2,3,5,11 In only two cases, including one of ours, was true spondylolisthesis demonstrated. Although it could be argued that the patient in our Case 1 did not require internal fixation and fusion, because the 3-month CT scan demonstrated callous formation, we were not willing to rely on that callous formation alone. Internal fixation and fusion with or without open reduction and/or decompression is the modern treatment paradigm for most thoracic fracture dislocations. 2,3,5,11 In the authors opinion, the primary goal is preservation of neurological function to allow normal activities of daily living including ambulation. The goal is not to have picture-perfect radiographic studies. As can be seen in Table 2, true anatomical reduction occurred in only FIG. 6. Case 3. Left: A CT myelogram obtained on postinjury Day 6, demonstrating a T-6 laminar fracture; the arrow indicates a bone fragment that telescoped beneath the lamina and compressed the spinal cord posteriorly. Right: A CT myelogram revealing an isodense epidural hematoma (arrow) compressing the spinal cord posteriorly. 336

Thoracic spinal fracture dislocation TABLE 1 Summary of findings in 15 cases in which there was traumatic thoracic fracture dislocation with minimal or no cord deficit Characteristic No. of Patients male 10 female 5 mean age (yrs) 34 mechanism of injury MVA car 6 (40) motorcycle 7 (47) plane crash 1 (6) fall from horse 1 (6) injury thoracic pain 15 (100) rib fracture(s) 14 (94) hemothorax 8 (53) anterior subluxation 6 (40) translation 15 (100) kyphosis 6 (40) bilat pedicle fractures 15 (100) treatment no op 5 (33) anterior approach 2 (13) posterior approach 5 (33) 360 op 3 (20) normal neurological outcome 14 (94) intercostal neuralgia 1 (6) one of 15 patients, and the same spinal deformity, despite normal/near normal neurological status, was present in six. The primary goal of surgery is restoration of spinal stability to prevent further subluxation and thus cord injury. To perform internal fixation and fusion via an anterior approach, posterior approach, or both remains a decision that must be tailored to the individual case. We recently reported that cadaveric fibula and Z-plate fixation following corpectomy for anterior and posterior column fractures, such as those in Case 4, can be performed with successful long-term results. 15 We made no effort to reduce the translation in any of the present four cases nor to undertake surgical reduction of any of the lesions in Case 1. Only in Cases 2 and 4 were canal decompressions performed as treatment for the mild neurological deficits. Even in the other 11 cases in which open reduction and internal fixation were performed, true anatomical alignment was reported in none. Interestingly, the only case in which true anatomical alignment outcome was documented was treated nonsurgically. Thus, if the patient suffers no deficits, then extreme attempts at surgical reduction may not be worth the risk of causing a neurological injury. TABLE 2 Summary of follow-up radiographic findings No Improved True No. of Change in Neurological Anatomical Treatment Patients Dislocation Function Alignment surgical* 10 3 7 0 nonsurgical 5 3 1 1 * Mean radiographic follow-up period 2.4 years (range 1 5 years). Mean radiographic follow-up period 1.2 years (0.3 3.2 years). Internal fixation and fusion, with or without spinal decompression depending on the degree of compression, is a reasonable alternative. References 1. Bohler L: Die Technik der Knochenbruchbehandlung im Frieden und im Kriege, vol 1, 9th 11th edns. Maudrich, Berlin, 1943 2. Bohlman H, Freehafer A, Dejak J: The results of treatment of acute injuries of the upper thoracic with paralysis. J Bone Joint Surg (Am) 67:360 369, 1985 3. Dekutoski M, Conlan S, Salciccioli G: Spinal mobility and deformity after Harrington rod stabilization and limited arthrodesis of thoracolumbar fractures. J Bone Joint Surg (Am) 75: 168 176, 1993 4. de Lucas J, Alvarez L, Abril J, et al: Fracture-dislocation of the thoracic spine without neurological lesion. Injury 25:105 107, 1994 5. Denis F: The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8: 817 831, 1983 6. Gertzbein S, Offierski C: Complete fracture-dislocation of the thoracic spine without spinal cord injury. A case report. J Bone Joint Surg (Am) 61:449 451, 1979 7. Hanley E, Eskay M: Thoracic spine fractures. Orthopedics 12:689 696, 1989 8. Harryman DT: Complete fracture-dislocation of the thoracic spine associated with spontaneous neurologic decompression. A case report. Clin Orthop 207:64 69, 1986 9. Korovessis P, Sidiropoulos P, Dimas A: Complete fracture-dislocation of the thoracic spine without neurologic deficit: case report. J Trauma 36:122 124, 1994 10. Liljenqvist U, Halm H, Castro W, Mommsen U: Thoracic fracture-dislocations without spinal cord injury: a case report and literature review. Europ Spine J 4:252 256, 1995 11. Magerl F, Aebi M, Gertzbein S, et al: A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3:184 201, 1994 12. Miyasaka Y, Satomi K, Sugihara S, et al: Posterior fracture-dislocation of the thoracic spine without neurologic deficit. A case report and short literature review. Spine 18:2351 2354, 1993 13. Roaf R: A study of the mechanics of spinal injuries. J Bone Joint Surg (Br) 42:810 823, 1960 14. Sasson A, Mozes G: Complete fracture-dislocation of the thoracic spine without deficit. A case report. Spine 12:67 70, 1987 15. Shapiro S, Bindal R, Abel T, et al: Cadaveric tibia and anterior Z-plate fixation after thoracic/lumbar corpectomy as compared to a posterior thoracic/lumbar vertebral body resection and posterior instrumentation. Neurosurgery 47:528 529, 2000 (abstr) 16. Simpson AH, Williamson DM, Golding SJ, et al: Thoracic spine translocation without cord injury. J Bone Joint Surg (Br) 72:80 83, 1990 17. Uriarte E, Elguezabal B, Tovio R: Fracture-dislocation of the thoracic spine without neurologic lesion. Clin Orthop 217: 261 262, 1987 18. Weber S, Sutherland G: An unusual rotational fracture-dislocation of the thoracic spine without neurologic sequelae internally fixed with a combined anterior and posterior approach. J Trauma 26:474 479, 1986 Manuscript received December 28, 2000. Accepted in final form November 13, 2001. Address reprint requests to: Scott Shapiro, M.D., Section of Neurosurgery, Room 323 East Outpatient Building, Wishard Hospital, 1001 West 10th Street, Indiana University Medical Center, Indianapolis, Indiana 46202. email: Sshapiro@iupui.edu. 337