Unstable thoracolumbar burst fracture usually requires a surgical treatment. In general, the goals of the surgery are to reduce the fracture, to correct the deformity, and to stabilize the spinal alignment until fracture healing is achieved. In the patients who encounter neurological deficit, the surgery should be focus to decompress the neural elements and also to prevent the progressive neural damage and facilitate neurological recovery by spinal fixation.1,2
Anterior direct decompression has been a recommended treatment for thoracolumbar burst fracture with neurological compromise.3,4 However, the surgery is much complicate and requires a special surgical expertise to operate.5 Moreover, it is necessary to remove considerable amount of bony fragments of anterior and middle column to achieve the adequate spinal clearance and need a strong anterior column supporting structure plus a rigid stabilization to enhance the successful outcome.3,4
On the contrary, posterior distraction and stabilization using pedicle screw instrumentation is less extensive and offers the comparable neurological outcome.6 Nevertheless, it is often difficult to achieve the efficient spinal canal decompression from posterior surgery alone, because the necessity of vertebral spinal column distraction to reduce the retropulsed bony fragments may jeopardize the stability of the posterior spinal fixation and result in late kyphotic deformity.7,8
Outcomes of the recent studies have demonstrated the mechanical advantage of the posterior spinal constructs when the pedicle screws are inserted at the level of fracture (index level) and also confirm the efficacy in the correction of spinal deformity, maintenance of correction, and preventing the failure of fixation in thoracolumbar fracture.9–11 In addition, we hypothesize that the insertion of index pedicle screws in combination with our design of the pedicle screw-plate system could facilitate the reduction of the collapsed vertebral body and restore the vertebral height which enhances the reduction of retropulsed fragments and achieves an adequate spinal canal decompression for neurological recovery without over distraction. Therefore, the aim of this study is to present an implement of our surgical techniques for the treatment of thoracolumbar burst fracture and to evaluate the surgical results.
MATERIALS AND METHODS
Thirty-one consecutive patients who sustained thoracolumbar burst fracture and had been treated with posterior spinal reduction and stabilization in our institute were included in this study. The surgical indications were as follows: (1) Spinal canal compromise more than 50%; (2) presence of kyphosis or scoliosis more than 20 degrees; (3) loss of vertebral height exceeding 50%; and (4) presence of neurological deficits. Patients with multiple spinal fractures were excluded from the study. This study was reviewed and had been approved by the hospital ethical research committee.
The radiographic evaluations were revealed before the operation, immediately after the operation and at the final follow-up using plain radiographs, anterior-posterior and lateral views. The bony union was also observed by plain radiographs. The kyphotic deformity was assessed on lateral radiograph by mean of Cobb measurement. The computed tomography scan was performed preoperatively to assess the degrees of the canal compromise at the injury level and verified the fracture configuration. The neurological assessments consisted of the evaluation of motor and sensory deficits and were stratified according to Frankel grading scale. The postoperative neurological outcomes were evaluated and compared with the preoperative status at the immediate postoperation and at the final follow-up.
Treatments and Operation Techniques
The intravenous methyl-prednisolone was prescribed for the patients with neurological deficit according to the NACIS III recommendation. Then, the operations were performed within the available time. The posterior stabilization was performed by using the pedicle screws plate (PSP) spinal system (T.K.S. Metal Works Co. Ltd, Bangkok, Thailand). The details of the implant designs and the operative techniques are as follows.
PSP System: the Components, Design, and Locking Mechanism
The PSP system consists of 4 main components; the screw, washer, nut, and spinal plate. The PSP spinal plate is a special design system for the dynamic correction of the deformity. It has 2 buttress square locking rods connecting with the small bridges and close ends. Our design of the spinal plate provides a slot for the screw shanks to engage between the 2 locking rods which is a fully constrained design. The washer is designed to constrain with the locking rods through the identical threads which force to securely lock to the PSP plate. The nut is used to push the washer down, along the proximal screw threads and to secure it against the PSP plate which is laid over the screw flare (Fig. 1). After achieving the secure locking mechanism of the PSP the remaining screw shanks, over the locked nuts and washers, can be easily cut off by further driven of the screws.
The patient was set in the prone position under general endotracheal anesthesia on the radiolucent table with the 2 dome-shaped cushions to support the sternum and the pelvis. Then the fracture level was identified by image intensifier. After the meticulous posterior surgical exposure, the pedicle screws were inserted at the vertebrae 1 level above and 1 level below the fracture site. On the fracture level, the pedicle screws are also inserted in every case which are so-called intermediate or index-level screws. We prefer to engage the screws at the fracture level underneath the upper vertebral endplate which the fracture pattern is commonly found, as burst fracture type B according to Denis et al13 classification. With our technique, all of the screws are partially engaged, not in their total depth. Usually, the proximal part of the screws are left 0.5 to 1 cm above the laminae (Fig. 2A). The final position of the screws were checked with image intensifier. The appropriate length of PSP plates (usually 70 mm long) were precontoured to the slight lordotic curve, which is 10 degrees in thoracolumbar region. Before the assembly of frame construct, the tip of screws on the vertebrae above and below the fracture level are relatively pointed in convergence because of kyphotic deformity of the fractured vertebra. When the precontour lordotic plates are applied, the screws are restrained into divergent positions, and the frame is constructed by the locking washers. Therefore, The distraction force is created at the fracture level because of the divergent position of the adjacent screws. Then, the preloaded frame construct is gradually driven into their depths by forward driven of the screws simultaneously, and the fracture site will be gradually distracted. The kyphosis deformity is spontaneously corrected as the fracture is reduced. The retropulsed fragment is subjected to be reduced by distraction of the fracture site by ligametotaxis of PLL. Finally, the vertebral height could be restored further with the secure locking of the frame construct (Figs. 2B, C). The over distraction is not allowed because of the chosen length of the PSP plates to control the height of the posterior column. Actually, slight compression of the screws along the PSP plates has to done if anterior overdistraction is detected by fluoroscopy during the gradual reduction. The laminectomy at the fracture level is performed if the patient has a neurological deficit or the vertical fracture of laminae is extensive and neural entrapment is observed. In case of neurological intact, a limited laminotomy is performed to prevent the buckle of ligamentum flavum into the spinal canal during the fracture reduction. No additional fusion is required.
The patients are allowed to be seated upright position in their beds 1 or 2 days after the surgery, depending on their general condition. After the drain was removed in the following day, the patients were encouraged to walk with gait support, no brace is necessary in general, except the patient with osteoporotic bone is advised to wear a thoracolumbar brace for preventing the progressive collapse. The patients are allowed to perform their normal activities of daily living without particular restriction depending on their capability and usually could return to work within a few months after the surgery. The instrument removal is not a routine of care. However, depend on the patient satisfaction, the implant removal will be allowed at least 1 year after the operation.
The statistical evaluation was performed with STATA version 8.0 software (TX). The 1-way analysis of variances with the post hoc analysis by Scheffe test was used to compare among the mean of the kyphotic angle before the operation, immediate postoperation, and the final results. The Fisher exact test was used for the quantitative data in the analysis of the change in Frankel score, and motor and sensory deficits. The significant level was set as P<0.05.
There were 14 males and 17 females with single level of thoracolumbar burst fracture included in this study. The mean age was 41 years (range, 16 to 79 y). The cause of injury included fall from a height and traffic accident. The injury levels were as follows: T11 (3 cases), T12 (1 case), L1 (15 cases), L2 (8 cases), L3 (3 cases), and L4 (1 case). The mean operative time was 126.7 minutes (range, 60 to 180 min) and the intraoperative blood loss was 454.8 mL (range, 150 to 2000 mL). The duration of follow-up was 1 year. The demographic data were illustrated in Table 1.
The Reduction of Kyphotic Deformity
The mean preoperative Cobb angle was −17.5±11.82 degrees. After the operation the kyphotic deformity was efficiently corrected. The mean Cobb angle of immediate postoperation was significantly reduced to 0.23±7.04 degrees (P<0.0001). At the final follow-up the mean kyphotic angle was reduced to −0.32±8.77 degrees; however, no statistically significant was demonstrated (P=0.974) when compared with the immediately postoperative results (Table 2). No nonunion of fracture or implant failure was observed in this study.
No neurological deterioration was found after the operation. The overall motor recovery was 9 out of 13 patients (69.2%). The motor power was shown an improvement immediately after the surgery in 1 patient; however, at the latest follow-up the improvement of motor power from grade 4 to grade 5 was found in 2 patients. In 4 patients, the motor power was improved from grade 3 to grade 5 and 2 patients grade 2 was improved to grade 4. Interestingly, 1 patient with complete paralysis, the motor power was recovered to grade 3 at the final follow-up and could walk with gait support (Table 3). According to Frankel score, the early improvement was found in 1 patient that was improved from Frankel D to E. The 5 patients who were in Frankel D had developed full neurological recovery (Frankel E) and 2 patients in Frankel C had improved to Frankel D. In 1 patient with complete neurological deficit had an improvement to Frankel D at final follow-up. No neurological recovery was found in 4 patients, in 3 patients remained in Frankel D, and 1 patient with complete neurological deficit was not shown any improvement (Table 4). The sensory improvement was demonstrated in 2 out of 7 patients (28.6%).
The posterior short-segment pedicle screws fixation, involving 1vertebra above and 1 below the fracture level, is now widely used for treating unstable burst fracture of the thoracolumbar spine. This technique is less extensive, simple, and offers the advantage of preserving lumbar motion segments. In addition, this technique provides rigid 3 columns fixation which is powerful enough to correct kyphotic deformity and to reduce the retropulsed bony fragments by the gradual reduction.8,14,15
Despite to these advantages, high implant failure rate and early loss of reduction have been reported.16,17 The large void created within the fractured vertebra after vertebral height restoration has been speculated as an important factor of these complications, which is mostly seen in short-segment fixation.11 The improvement of biomechanical stability by adjustment of the pedicle screws has been reported by Oda et al.7,18 The results demonstrate that with distraction and extension of the instrument produces the closest anatomic reduction. Unfortunately, with this adjustment, the stability of the implant construct may jeopardize as it provides less stability to sustain the compression load when compared with the compression adjustment instrumentation.
In attempt to prevent early failure of the short-segment pedicle screw fixation, the alternative techniques to augment the fractured vertebra have been advocated for instance, transpedicular grafting,19,20 or injection with various kinds of cement or derivatives.20–22 Another option is insertion of the pedicle screws into the fractured vertebra level (index-level screw). These screws should provide a buttress effect to support the upper vertebral end plate and may support anterior column of the injury level which should maintain the stability of the frame construct. These advantages have been confirmed both in clinical and biomechanical studies.9–11
In this study, we described our technique for the reduction of the burst fracture. With the precontour lordotic plate, we can achieve the nearly anatomic reduction by extension and distraction of the fractured vertebra as pointed out by Oda et al18 and Steib et al.23 However, with our PSP plate design in which, the upper and lower screws are constrain locked within the length of PSP plate, the over distraction during the operation which may lead to early failure can be prevented (Fig. 3). Moreover, with the insertion of index-level screws, the fractured void could be partially replaced and the gradual driven of the index-level screws can facilitate the reduction of the retropulsed bony fragments by pushing them into the void (Fig. 4). In addition, the index pedicle screws should provide a buttress effect to prevent the vertebral end plate collapse after reduction and support the anterior column of the injury level. They are also constrain locked to the PSP plate in which they should maintain the stability of the implant construct as previously described.
The results of our study have shown that, the kyphotic deformity is efficiently corrected after the operation. The mean Cobb angle at immediate postoperation is significantly reduced and could maintain until the healing process established. However, at the final follow-up some degrees of the reduction loss are demonstrated as slight increase in kyphotic deformity, but no statistically significant is demonstrated when compared with the immediate postoperative results (Table 4). These results are comparable with previous study by Guven et al.11 However, in our experience, the screw withdrawn may be observed in the lowest pedicle screws in senile osteoporotic patients.
Kong et al2 have introduced the technique of direct reduction of the retropulsed bone fragment from the posterior laminectomy. However, this technique required special self-design instrument and needed experiences of the surgeon to do the operation. The results for this study revealed that 77% of the patient had neurological improvement. In this present study, we used an alternative approach for the treatment. With our technique, we performed simple laminectomy at the fracture level, in combination with the vertebral body height restoration by insertion of the index pedicle screws, preload frame construct in an extended position, and gradual loading the frame into involved vertebrae which should also facilitate the spinal canal clearance by gradual reduction of the retropulsed fragment into the void. The overall neurological improvement in our series is 70% that should be comparable with the previous studies.1,24 The increase in kyphotic deformity after decompression and fixation in neurologically compromised patients was not demonstrated in our study (Fig. 5).
In this present study, we did not evaluate the quantity of the reduction of the canal encroachment. However, there are conflicting evidences regarding the degrees of canal encroachment and neurological deficit.1 In addition, there are still lack of strong evidences to support the necessity of complete reduction or removal of the retropulsed bony fragments to promote the neurological recovery.1,25,26 Furthermore, the spontaneous remodeling of the retropulsed fragment is well established.1,27 Thus, we do convince that our surgical technique could reduce the retropulsed fragment in some extent and the remaining parts should be further spontaneously remodeled to the normal configuration that result to the improvement of neurological deficit in this present study.
The result from our present study confirmed that the short-segmental pedicle screws instrumentation with index-level screws can achieve the strong implant construction for the reduction and maintain kyphosis deformity correction. Our technique of fixation to control posterior column height and gradual fracture reduction could provide better restoration of the vertebral body height and could facilitate the reduction of the retropulsed bony fragments. In addition, the simple laminectomy or laminotomy could be performed without further increase of the kyphotic deformity. However, the retropulsed bony fragments may not be completely reduced with our present technique and need further study to confirm the quantity of the reduction of the canal encroachment and the potential of spinal canal remodeling.
1. Rath SR, Kahamba JF, Kretschmer T, et al..Neurological recovery and its influencing factors in thoracic and lumbar spine fractures after surgical decompression and stabilization.Neurosurg Rev.2005;28:44–52.
2. Kong W, Sun Y, Hu J, et al..Modified posterior decompression for the management of thoracolumbar burst fractures with canal encroachment.J Spinal Disord Tech.2010;23:302–309.
3. Miyakoshi N, Abe E, Shimada Y, et al..Anterior decompression with single segmental spinal interbody fusion for lumbar burst fracture.Spine.1999;24:67–73.
4. Shi R, Liu H, Zhao X, et al..Anterior single segmental decompression and fixation for Denis B type thoracolumbar burst fracture with neurological deficiency: thirty-four cases with average twenty-six month follow-up.Spine.2011;36:E598–E605.
5. Verlann JJ, Diekerhof CH, Buskens E, et al..Surgical treatment of traumatic fractures of the thoracic and lumbar spine: a systematic reviews of the literature on techniques, complications and outcome.Spine.2004;29:803–814.
6. Sjostrom L, Karlstrom G, Peeh P, et al..Indirect spinal canal decompression in burst fractures treated with pedicle screw instrumentation.Spine.1996;21:113–123.
7. Oda T, Panjabi MM.Pedicle screw adjustments affect stability of thoracolumbar burst fracture.Spine.2001;26:2328–2333.
8. Andress HJ, Braun H, Helmberger T, et al..Long-term results after posterior fixation of thoraco-lumbar burst fractures.Injury.2002;33:357–365.
9. Mahar A, Kim C, Wedemeyer M, et al..Short-segment fixation of lumbar burst fracture using pedicle fixation at the level of fracture.Spine.2007;32:1503–1507.
10. Baaj AA, Reyes PM, Yaqoobi AS, et al..Biomechanical advantage of the index-level pedicle screw in unstable thoracolumbar junction fractures.J Neurosurg Spine.2011;14:192–197.
11. Guven O, Kocaoglu B, Bezer M, et al..The use of screw at the fracture level in the treatment of thoracolumbar burst fractures.J Spinal Disord Tech.2009;22:417–421.
12. Chanplakorn P, Wajanavisit W, Laohacharoensombat W.Apical derotation for deformity correction in aldolescent idiopathic scoliosis using the pedicle screws-plate spinal system: surgical technique and results.Tech Orthop.2011;26:203–211.
13. Denis F.The three-column spine and its significance in the classification of acute thoracolumbar spinal injuries.Spine.1983;8:817–831.
14. Dai LY, Jiang LS, Jiang SD.Posterior short-segment fixation with or without fusion for thoracolumbar burst fractures. A five to seven-year prospective randomized study.J Bone Joint Surg Am.2009;91:1033–1041.
15. Zou D, Yoo JU, Edwards WT, et al..Mechanic of anatomic reduction of thoracolumbar burst fracture: comparison of distraction versus distraction plus lordosis, in the anatomic reduction of the thoracolumbar burst fracture.Spine.1993;18:195–203.
16. McLain RF, Sparling E, Benson DR.Early failure of short-segment pedicle instrumentation for thoracolumbar burst fractures. A preliminary report.J Bone Joint Surg Am.1993;75:162–167.
17. Muller U, Berlemann U, Sledge J, et al..Treatment of thoracolumbar burst fractures without neurological deficit by indirect reduction and posterior instrumentation: bisegmental stabilization with monosegment fusion.Eur Spine J.1999;8:284–289.
18. Oda T, Panjabi MM, Kato Y.The effects of pedicle screw adjustments on the anatomical reduction of thoracolumbar burst fractures.Eur Spine J.2001;10:505–511.
19. Alanay A, Acaroglu E, Yazici M, et al..Short-segment pedicle instrumentation of thoracolumbar burst fractures: does transpedicular intracorporeal grafting prevent early failure?Spine.2001;26:213–217.
20. Liao JC, Fan KF, Kaorochana G, et al..Transpedicular grafting after short-segment pedicle instrumentation for thoracolumbar burst fracture: calcium sulphate cement versus autogenous iliac bone graft.Spine.2010;35:1482–1488.
21. Cho DY, Lee WY, Sheu PC.Treatment of thoracolumbar burst fractures with polymethyl methacrylate vertebroplasty and short-segment pedicle screw fixation.Neurosurgery.2003;53:1354–1360.
22. Fuentes S, Blondel B, Metellus P, et al..Percutaneous kyphoplasty and pedicle screw fixation for the management of thoraco-lumbar burst fractures.Eur Spine J.2010;19:1281–1287.
23. Steib JP, Charles YP, Aoui M.In situ contouring technique in the treatment of thoracolumbar fractures.Eur Spine J.2010;19suppl 1S66–S68.
24. Schnee CL, Ansell LV.Selection criteria and outcome of operative approaches for thoracolumbar burst fractures with and without neurological deficit.J Neurosurg.1997;86:48–55.
25. Pait C, Djian P, Goutallier D.Factors influencing neurological recovery after corporectomy for thoracolumbar spine fracture.J Bone Joint Surg Br.1993;75suppl180.
26. Herndon WA, Galloway D.Neurologic return versus cross-sectional canal area in incomplete thoracolumbar spinal cord injuries.J Trauma.1988;28:680–683.
27. Dai LY.Remodelling of the spinal canal after thoracolumbar burst fractures.Clin Orthop.2001;382:119–123.
Keywords:© 2014 by Lippincott Williams & Wilkins
thoracolumbar fracture; short-segments fixation; surgical technique; lordosis fixation; pedicle screws fixation