Burst fracture of the fifth lumbar vertebra: results of posterior internal fixation and transpedicular bone grafting

Transcription

1 Eur Spine J (2002) 11 : DOI /s ORIGINAL ARTICLE A. Kaminski E. J. Müller G. Muhr Burst fracture of the fifth lumbar vertebra: results of posterior internal fixation and transpedicular bone grafting Received: 28 June 2001 Revised: 1 December 2001 Accepted: 11 January 2002 Published online: 29 August 2002 Springer-Verlag 2002 A. Kaminski ( ) E.J. Müller G. Muhr Chirurgische Klinik und Poliklinik, Berufsgenossenschaftliche Kliniken Bergmannsheil, Universitätsklinik, Bürkle-de-la-Camp-Platz 1, Bochum, Germany akaminski@t-online.de, Tel.: , Fax: Abstract Burst fractures of the fifth lumbar vertebra are rare, and there are only a few reports on this subject, which is characterised by its unique anatomical and biomechanical features. This retrospective analysis reports on ten patients whose fractures of L5 were stabilised with a short internal fixator in combination with a posterior fusion as well as transpedicular bone grafting. The average follow-up period was 22 months. Radiometric data were surveyed and compared to the functional results. The height of the fractured vertebra remained nearly unchanged throughout the course. Loss of lordosis of 4 in the upper disc space and 4 in the lower disc space were observed postoperatively. At the time of follow-up, the values of segmental lumbar lordosis were significantly below the preoperative level. The narrowing of the neural canal was reduced from 57% to 28% with the surgical intervention. However, there was no correlation between the functional and the radiological outcome. Neurological deficits were documented in two patients, which declined during the course. There were no severe postoperative complications. The results of the present study demonstrate that the described surgical procedure in fractures of the fifth lumbar vertebra does not promote an anatomic restoration of the fractured vertebra, nor of the segmental lordosis. However, the clinical results do not correlate with the radiological outcome. Nonoperative treatment with early mobilisation without external support seems to be the treatment of choice. Keywords Burst fracture Lumbar spine Fifth lumbar vertebra Management Introduction Burst fractures of the thoracolumbar spine are the result of an axial compressive force, a concept that was described by Holdsworth in 1963 [10] for a two-column classification system, as well as by Denis, who modified this theory with the introduction of the third, middle, column into the classification system [6]. Denis also emphasised that burst fractures are inherently unstable [6]. Burst fractures of the fifth lumbar vertebra are very rare. The unique anatomical and biomechanical characteristics of this region are the shape of the vertebral body, the position below the apex of normal lumbar lordosis, the increased ranges of motion in the lumbosacral articulation, and the associated ligaments between L5 and the pelvic ring. The neural canal is widest in this area and the neurologic dysfunction results from injury to the cauda equina, without involvement of the conus medullaris. To our knowledge, only four papers considering the peculiarities of L5 fractures and evaluating the results of different treatment protocols have been published [5, 7, 9, 13]. Mick et al. [13] reported on 11 cases and on the results of treatments based on two different concepts. This article

2 436 Fig. 1 Fracture classification according to Magerl et al. [12] reviews the results of ten burst fractures of L5 treated with a single operative method posterior internal fixation, posterolateral fusion and transpedicular bone grafting with regard to the functional and radiographic outcome. Materials and methods Between 1993 and 1997, ten patients, three female and seven male with an average age of 29 years (range: years), were treated operatively following an unstable burst fracture of the fifth lumbar vertebra. The fractures were classified according to Magerl et al. [12] (Fig. 1). In all patients, treatment consisted of application of a bisegmental internal fixator (USS, Synthes Co.), posterolateral fusion as well as transpedicular intervertebral bone grafting. The duration of the surgical procedure was 161 min on average (range: min). In five patients (50%) a laminectomy was performed, because of neurological deficits at the time of admission in two of these (20%) graded as Frankel C and D respectively [8] and in three patients (30%) because of significant narrowing of the neural canal. In five patients (50%) the injury was due to a fall from height, three patients (30%) had been injured in a road traffic accident and two patients (20%) had been hit by an object. In 50% of the patients, accompanying injuries were documented (3.2 injuries on average; see Table 1). All patients were followed up, at 22 months on average (range: months). The radiometric evaluation was performed on anteroposterior and lateral radiographs of lumbar spine according to the Cobb method [4]. The determined parameters are defined in Fig. 2 and Table 2. The α-angle represents the extent of the vertebral wedging at the fracture level, whereas the β-angle reflects the changes in the adjacent cranial, and the γ-angle the changes in the caudal, intervertebral disc space. The evaluation of the pre- and postoperative narrowing of the neural canal in the sagittal projection, the spinal compression (SC), was based on computer tomographic (CT) scans. In addition to the objective radiometric data, the subjective result of the patients was evaluated using a questionnaire to establish the pain and the activity status. Based on these data, a classification of the results into four groups excel- Table 1 Associated injuries (n=16 in five patients) Lumbar spine L1 fracture, stable 1 L2 fracture, stable 1 L3 fracture, stable 1 L4 fracture, stable 2 Central nervous system Commotio cerebri 1 Pelvis Isolated fracture of the superior pubic ramus 1 Long bones Femur fracture 2 Tibia fracture 3 Thorax Rib fractures 2 Other lesions Calcaneus fracture 2 Fig. 2 Measuring technique

3 437 Table 2 Definition of the determined radiological parameters α-angle β-angle γ-angle Scoliosis Angle between the inferior and the superior end plate of the fractured vertebral body (lateral) Angle between the inferior end plate of the fractured vertebral body and the superior end plate of the vertebral body above the fracture level (lateral) Angle between the superior end plates of the vertebral bodies above and below the fracture level (lateral) Angle between the inferior end plate of the fractured vertebral body and the superior end plate of the vertebral body above the fracture level (AP) lent, good, fair and poor was conducted. A detailed description of the individual groups can be found in Table 3. Statistical analysis was performed using the t-test, with a significance level P<0.05. The value characteristics (normal distribution) were determined with the Kolmogorov-Smirnov-test. Results There were no intraoperative and no neurological complications. A postoperative haematoma had to be evacuated in one patient (10%). After revision, the further postoperative course was uneventful. Regarding the radiographic parameters, the initial α-angle averaged 0, which corresponds to only a moderate loss of height. With the operative intervention, the value decreased to 1, which was not statistically significant (P=0.52), and remained unchanged during the further course. The alteration of the segmental lordosis (β-angle), which was 13 preoperatively and 12 after the dorsal instrumentation, was not significant either (P=0.859) (Fig. 3). However, at follow-up, a significant loss of segmental lordosis of 4 (from 12 to 8 ) was observed (P=0.008). The same behaviour was documented for the γ-angle, with a loss of 8 (from 28 to 20 ) post- Table 3 Pain and activity findings at the time of followup Result Pain status Activity status Excellent No pain Unlimited return to previous activities (employment and activities of daily living) Good Occasional or mild pain, no need Return to previous activities (employment and for analgesics activities of daily living) with restrictions Fair Moderate pain, occasional pain No return to previous activities (employment medication and activities of daily living) modified work and activities Poor Severe or constant pain, requiring No return to work, totally disabled continual pain medication Fig. 3 Lateral radiograph of an L5 burst fracture. Note the maintenance of alignment of the α-angle and the significant changes with loss of lordosis of the β- and γ- angle over the course

4 438 Table 4 Radiographic parameters on admission, following surgery and at follow-up (negative values indicate lordosis, positive values indicate kyphosis) Preop. mean (range) Postop. mean (range) Follow-up mean (range) α-angle 0 ( 10 to 15 ) 1 ( 6 to 3 ) 1 ( 13 to 5 ) β-angle 13 ( 25 to 12 ) 12 ( 21 to -4 ) 8 ( 16 to 2 ) γ-angle 29 ( 51 to 4 ) 28 ( 34 to 21 ) 20 ( 32 to 12 ) Scoliosis 5 (0 to 15 ) 1 (0 to 7 ) 1 (0 to 7 ) Spinal compression 57% (10% to 87%) 28% (7% to 62%) Fig.4 Summary of radiographic data with their correlation coefficients operatively (P=0.008). The posttraumatic scoliotic deformity of 5 decreased significantly, to 1, with posterior instrumentation (P=0.01). This value remained constant during the course. The radiographic parameters are summarised in Table 4 and Fig.4. The mean preoperative spinal compression was 57%, which was reduced to 28% with the operative intervention. In two patients with neurological compromise, a preoperative spinal compression of 86% and 60% respectively was documented. In the remaining neurologically intact patients, the spinal compression was 50% on average. However, there was no positive correlation between the spinal compression and the occurrence of neurological deficits (Spearman-ρ correlation coefficient = 0.479, P=0.277) (see also Fig. 5). In both patients, an improvement of the neurological function by one Frankel grade was observed. Regarding the pain status at follow-up, four patients (40%) had an excellent result, three patients (30%) were rated as good, and in a further three patients (30%) the result was fair. A similar distribution was documented for the activity level, with five (50%) excellent, two (20%) good and three (30%) fair results. There were no poor results. Statistical analysis did not show a positive correlation between the pain or the activity status and the documented radiological features (P=0.607 and P=0.110, respectively; correlation test according to Spearman-ρ). Discussion Fractures of the lumbosacral spine have to be distinguished from fractures of the thoracolumbar spine, because of the particular anatomic and biomechanical characteristics of this region of the spine. Only a few reports have focused on fractures of the fifth lumbar vertebra, and the results and recommendations for treatment are controversial. Finn and Stauffer [7] report on seven patients with a fracture of the fifth lumbar vertebra, who were treated nonoperatively by an elaborate resting position in a bodyjacket. For the description of the radiological characteristics, a parameter corresponding to the γ-angle was chosen [7]. The initial posttraumatic value of 29 corresponds

5 439 Fig. 5 This 45-year-old woman was involved in a motor vehicle accident, sustaining an L5 burst fracture. Results from her neurologic examination were normal closely to our results, and did not change significantly with nonoperative treatment. The height of the vertebral body also remained nearly the same [7], as was the case in our series. In contrast to the results of Finn and Stauffer [7], Mick and co-workers [13] described a significant loss of height in the anterior column in five patients treated nonoperatively, whereas in six fractures that underwent posterior stabilisation, an increase of height in the anterior column was noted. However, the authors did not comment on changes in the adjacent intervertebral spaces [13]. This is in contrast to our findings, where a significant loss of lordosis was documented, despite rigid dorsal stabilisation and additional intervertebral bone grafting. An explanation for this observation could be that, with posterior instrumentation, the height of the anterior column was not restored, which is reflected in the unchanged lordosis angle after surgery, but the posterior column was overdistracted. This results in a secondary collapse of the anterior column in the intervertebral disc spaces, with further loss of lordosis. However, despite the significant loss of lordosis, the functional outcome did not show any correlation with the radiological parameters an observation that has been made by other authors as well [5, 7]. In the series of Mick et al., the majority of fractures treated nonoperatively showed excellent functional results, despite the significant loss of height of the anterior column. These authors could not find any correlation between the radiological parameters and the functional outcome either [13]. In view of the reported data in the literature and our own results, posterior segmental stabilisation of burst fractures of the fifth lumbar vertebra cannot be recommended on a routine basis. The question of whether circumferential stabilisation provides better radiological results has yet to be answered, but obviously the radiological result does not correlate with the functional outcome. It has yet to be shown that better radiological results correlate with an improved clinical outcome. Therefore, nonoperative treatment with early mobilisation without external support seems to be the treatment of choice. Operative stabilisation is only recommended if pain during mobilisation is not tolerated despite adequate pain medication. Particular attention must be paid to fractures of the spine with neurological deficits. However, in this region it is no longer the spinal cord that is affected. The nerve roots that form the cauda equina originate from the second motoneuron, and neurological deficits at this level are similar to peripheral nerve lesions and demonstrate a prognostically advantageous tendency for spontaneous recovery [2]. The cauda equina shows a higher resistance than the conus medullaris to external effects [1]. A correlation between the severity of the neurological deficits and the extent of neural canal narrowing could not be proven in the present study an observation that correlates well with the report of Finn and Stauffer [7]. Boerger et al. [3], after a metaanalysis of the available literature, concluded that the neurological injury occurs exactly at the time of the impact and is, therefore, not represented by the geometric parameters at the time of the survey of the findings. For example, the pressure within the spinal canal at the moment of the impact is higher than immediately after the trauma [11]. Consequently there are significant doubts that improvement of the neurological situation can be achieved by surgical clearance of the canal. Interestingly, there is a high coincidence between fractures of the fifth lumbar vertebra and the occurrence of further, stable fractures in the remaining lumbar spine. Statistically, every second unstable fracture of the fifth lumbar vertebra is accompanied by a stable fracture of the first to fourth lumbar vertebra. In comparison, accompanying injuries of the upper extremities have not been documented.

6 440 Conclusions The results of the AO internal fixator technique does not significantly modify the radiological features of the burst fractures of the fifth lumbar vertebra. In the absence of scientific evidence of the neurological benefits of canal clearance, the use of this technique can not be routinely recommended. References 1. An HS, Vaccaro A, Cotler JM, Lin S (1991) Low lumbar burst fractures. Comparison among body cast, Harrington rod, Luque rod, and Steffee plate. Spine [16 Suppl]: Andreychik DA, Alander DH, Senica KM, Stauffer ES (1996) Burst fractures of the second through fifth lumbar vertebrae. J Bone Joint Surg Am 78: Boerger TO, Limb D, Dickson RA (2000) Does canal clearance affect neurological outcome after thoracolumbar burst fractures? J Bone Joint Surg Br 82: Cobb JR (1948) Outline for the study of scoliosis. In: American Academy of Orthopedic Surgeons: Instructional Course Letters, no. 5. Edwards Bros, Ann Arbor, p Court-Brown CM, Gertzbein SD (1987) The management of burst fractures of the fifth lumbar vertebra. Spine 12: Denis F (1983) The three column spine and its significance in the classification of acute thoracolumbar spine injuries. Spine 8: Finn CA, Stauffer ES (1992) Burst fracture of the fifth lumbar vertebra. J Bone Joint Surg Am 74: Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS, Ungar GH, Vernon JDS, Walsh JJ (1969) The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia 7: Fredrickson BE, Yuan HA, Miller H (1982) Burst fractures of the fifth lumbar vertebra: a report of four cases. J Bone Joint Surg Am 64: Holdsworth FW (1963) Fractures, dislocations and fracture-dislocations of the spine. J Bone Joint Surg Am 54: Limb D, Shaw DL, Dickson RA (1995) Neurological injury in thoracolumbar burst fractures. J Bone Joint Surg Br 77: Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3: Mick CA, Carl A, Sachs B, Hresko T, Pfeifer BA (1993) Burst fractures of the fifth lumbar vertebra. Spine 18: