Combined anterior and posterior decompression and short segment fixation for unstable burst fractures in the dorso lumbar region.
The dorso lumbar segment of spine (D10 to L2) is an unstable zone between fixed dorsal and mobile lumbar spine. A combined anterior and posterior approach with short segment stabilization was found most appropriate. Thirty cases were treated over a period of 4 years and 6 months. There were 26 male and 4 female patients with mean age of 32.6 years. L1 vertebra was fractured in 17 cases, D1 in 8 cases, D11 in 4 cases and D10 in one case. 14 cases had total neurological deficit, 9 cases had partial and 7 had no neurological deficit. We have used three column classification of Denis to assess the cases. Seven patients returned to regular physical work, 5 had restricted physical work, 5 remained in full time light job and 9 patients were unable to return to original job but did some work. Most had flaccid paraplegia but 4 patients were completely disabled due to spastic paraplegia. Neurological recovery occurred in all the patients with partial paralysis, and appeared to be dependent on initial kyphosis. The overall recovery rate varied from 50% to 90%. There is no correlation between canal compromise and severity of injury. Neurological injury occured at the time of trauma, rather than as a result of pressure of fragment in the canal. No strong conclusion could be drawn to say that the results of surgery were superior to non-operative treatment.
The dorso lumbar (DL) segment of spine (D10 to L2) is an unstable zone between fixed dorsal and mobile lumbar spine at a junction of dorsal kyphosis and lumbar lordosis. Acute injury to DL segment is the second most frequent site after cervical spine in adults. The injury, although not associated with high mortality, causes severe morbidity (mortality 0.5% as compared to 20% in the cervical spine). In India, majority of patients have axial load injury with unstable burst fractures of the vertebral bodies. It is estimated that approximately 75% of patients with DL injuries sustain some degree of neurological deficit. Treatment of DL fractures constitute a most controversial subject. There are strong proponents of conservative therapy,,,,,, and those that believe in surgical management.,,,,,,, Over a period of twenty years, following a trial of various methods of posterior and anterior decompression and stabilization using, both long and short segment constructs, the authors feel that with the availability of locally manufactured instruments and implants in our country, a combined anterior and posterior approach with short segment stabilization was most appropriate. A series of 30 cases thus treated over a period of 4.5 years from June 1994 till December 1998 have been reviewed.
The main aim has been to achieve early and solid spinal stability at the site of disruption. Ignoring the controversy, seven cases with unstable burst fractures without neurological deficit were treated by this surgical method.
Thirty cases treated with combined anterior and posterior approach with short segment stabilization over a period of 4.5 years were reviewed. There were 26 male and 4 female patients. Age ranged from 18 to 51 years (mean 32.6 years). L1 vertebra was fractured in 17 cases, D12 in 8 cases, D11 in 4 cases and D10 in one case. In 3 cases of L1 fracture, L2 body was partially involved. Fourteen cases had total neurological deficit, 9 had partial neurological deficit and 7 had no neurological deficit. Majority of patients came late from peripheral hospitals, after receiving conservative treatment for a varying period. More than 75% of cases were admitted after 72 hours.
Stability and Instability
We used 3 column classification of Denis to assess the cases [table I]. In unstable fractures the bony spinal canal is compromised to a varying degree and, in some of them, there is neurological deficit. Punjabi supported the three column theory of Denis and bolstered the concept of middle column (of which posterior longitudinal ligament is a component) being the primary determinant of mechanical stability of dorso lumbar region of the spine. Punjabi also described criteria to label a burst fracture unstable [table II]. We used the following criteria to call a burst fracture unstable in our series: i) severe and persistent pain with slightest movement, ii) more than 20o kyphosis deformity, iii) more than 50% anterior vertebral body compression, iv) canal compromise of 40% from T11 to L1, v) associated injury to posterior column, vi) association of neurological deficit.
The compromise of bony spinal canal
It is difficult to clarify a threshold that needs surgical decompression to prevent late neurological deficit in dorso lumbar burst fractures. Crompinger set a line at 50% and suggested that more than 50% canal compromise should be decompressed surgically. Knight gave more stress on kyphosis associated with burst fractures with canal compromise and felt that rather than measuring the canal compromise, the patient should be considered for surgery in case kyphosis is more than 20 degrees. Hashimoto et al showed significant ratios at which neurological injury became probable and therefore should be considered for surgery. They felt that canal compromise of 35% at T11-T12, 45% at L1 and 55% at L2 should be considered for surgery.
All patients were operated upon within two weeks. A set protocol was established. The injury was assessed. Necessary investigations were carried out. In the first stage of surgery the spine was approached posteriorly in prone position. Decompression of the cord at the site of injury was carried out. Short segment stabilization was then done with one vertebra above and one vertebra below, using 4.5mm x 36 mm pedicle screws and VSP plates manufactured locally. The spine was stabilized in neutral position. Four days later, as a second stage, thoracotomy was done through the bed of 9th rib on the left side in lateral position. Left lung was deflated by blocking the left bronchus and the spine was approached anterolaterally as described originally by Hodgson. The fractured body was radiologically identified. Posterior 2/3rd or the whole body was then excised to decompress the dura. Reconstruction was carried out by using left iliac crest autograft. The fusion was further strengthened by using a strong reconstruction plate and preferably bicortical purchase screws, one in each vertebra above and below [Figure 1a] and [figure b]. By this method adequate decompression and satisfactory stabilization was carried out anteriorly and the deficit in the strength of the construct was made good by posterior pedicle screws stabilizing construct. The anterior horizontally placed screws in the vertebral bodies made good for the absence of cross bars posteriorly. Those patients with partial neurological deficit or no neurological deficit were quickly mobilized within a week with external orthosis, which was maintained for a minimum period of 4 months. If there was pain during movements, the orthosis was maintained for a further period of 2 months.
The results were assessed using Denis scales.
Work Scale : Sixteen patients had partial or no neurological deficit. Five of these were students of young age and were pursuing their studies. Six patients belonged to heavy labour group and only two of these went back to work; but were doing regular but relatively lighter duties. Four patients could not return to their jobs. Of the remaining five patients, one female patient could not return to her job but was happy doing household work, and four patients (2 working in mills and 2 in the office) returned to their original sedentary duties. Systemic pharmacological agents (baclofen, tizanidine, diazepam) did not produce the desired benefit in relieving hypertonia on a long term basis. The results of Denis Work Scale are shown in [Table III]
Neurological improvement : All nine patients with partial neurological deficit in this series, showed significant improvement. Two students out of five were using ankle covering boots to cope up with residual weakness in the ankle joint. Three students were using normal shoes. The improvement using Frankel's grades is shown in the [Figure - 2]. Assessment was done at the end of eight months. The muscle showing maximum deficit has been taken as the grade point on Frankel's scale. On the whole, most patients were happy with surgery as far as relief of pain was concerned. Most patients did not require medication for pain. Pain, thus, was not a significant problem in these cases. With right internal dual stabilization and with external orthosis, good stability was achieved.
Complications : One patient died due to pneumonia while under rehabilitation, one month after injury. In several patients lateral cutaneous nerve was damaged while taking graft from the iliac crest. This caused paresthesia in the thigh in those having sensations in the limbs. No special treatment was given. Bed sores in paraplegic patients is an accepted reality in our country and on number of occasions the help from plastic surgeons was required to do a rotation flap. The rate of infection was high (33%), but it was contained. The infection was not serious enough to necessitate removal of implants. So far the implants have not been removed in any patient. One student developed severe epidedymo-orchitis due to self catheterization 15 months after surgery. Bladder stones were removed from three patients. Calcium was given to every patient to achieve good bony arthrodesis. Possibly it had been responsible for the formation of bladder stones. Dural laceration with or without CSF leak occurred in seven cases in this series. It should be suspected when posterior column is injured and particularly when there is fracture of lamina.,
Neurological injury and recovery : In this series neurological recovery occurred in all patients with partial paralysis. Neurological recovery appeared to be dependent on initial kyphosis and 90% recovery was expected in patients with greater than 15 degree kyphosis. The series recovery period was taken as 8 months, when maximum recovery should have occured, although some recovery may still be seen after one year. Vertical fractures with maximum canal compromise, impacting the neural elements against posterior bony arch, caused greatest neurological damage with less favourable recovery pattern in our cases. The overall recovery rate in our patients varied from 50 to 90%.
The management of injuries at dorso lumbar junction has remained controversial. The problem of making decisions about therapy seems to lie in identifying patients separately for whom operative or non operative treatment would yield the most satisfactory results. Short segment fixation was found to be more useful in our patients. It uses the simple principles of distraction or compression or cantilever bending construct. Long posterior rodding uses two biomechanical principles of : i) intermediate points of fixation and ii) the use of three point bending construct that uses a fulcrum for stabilization. This can cause hook or sublaminar wire fatigue. Short segment fixation results in less spinal stiffness. Forces applied to the spine are not strong and fatigue failure is uncommon.,, It is not necessary to remove the construct, although this is necessary with long rods. However, admittedly, short segment stabilization has limitations with translational injuries where better success can be achieved with long rods stabilization. Short segment posterior pedicle screw fixation technique to resist axial spinal loading anteriorly is not adequate. Sasso, in fact, has established that only posterior stabilization either with Harrington rods, Luque rods or pedicle screws and plates failed to maintain sagittal correction at 12 months. McLain et al found high rate of failure with posterior fixation alone with short segment CD instrumentation. But if the construct is used in neutral mode and adequate strut support is provided anteriorly, the efficacy and utility of pedicle screws is increased. Adequate anterior structural support was provided in our cases by putting a solid iliac crest strut graft and it was further strengthened with metallic implants using reconstruction plate and screws. The biomechanics of this construct is acceptable.
Kaneda's device began the modern era of short segment anterior spinal stabilization. With plates hooked to vertebral bodies, two rods for stabilization and two bicortical purchase screws are placed in each vertebral body above and below the posterior stabilization. In the absense of easy accessibility to Kaneda implants, the present construct using the principles of short segment stabilization does appear most satisfactory. It achieves solid interbody fusion without causing significant kyphotic deformity.
In this series, seven cases without neurological deficit were operated upon. The main aim had been : i) anatomical reduction, ii) rigid stability, iii) early rehabilitation and iv) return to work and gainful employment. Treating burst fractures in dorso lumbar region with or without neurological deficit has remained controvercial. Good long term results of non operative treatment i.e. less backpain, satisfactory work status, no neurological deterioration and no reported.,,,,, complications have been Advocates of conservative treatment are satisfied that spinal canal compromise even greater than 50% gets resorbed spontaneously within one year. The possibility of neurological deficit occurring late in patients treated conservatively cannot be denied. Denis et al reported that 17% of patients developed late neurological deficit, when there was no neurological deficit initially. These patients required surgery for decompression. Bohiman et al established that late onset neurological deficit and pain in conservatively managed patients can be satisfactorily treated with decompression. Transfeldt et al found 68% improvement in neurological deficit when operated within a period of two years and 43% improvement when operated after two years. Late onset neurological deficit is thus a reality in these patients and that decompression does help to relieve pain and improve neurological deficit. Some patients treated conservatively following decompression fractures without neurological deficit have returned back with progressive kyphosis with pain and or neurological deficit. These patients require stabilization following compression of the dorsolumbar junction. It is possible that these patients had instability from the time of injury and could not take the axial load on mobilization and developed kyphosis. Whitesides and Sutherland et al reported relief of pain and improvement in neurological deficit following correction of kyphosis even after two years. The major benefit, however, lies in relief of pain.
Since 1989, the trend has been towards short period of bed rest in most centers. Weinstein et al found a similar trend towards shorter hospitalization in patients treated in recent times in his retrospective study. Four weeks bed rest and external orthosis for 12 weeks is now accepted as standard conservative treatment.
Advocates of surgery believe in obtaining early and solid spinal fusion and good stability. Excellent results with operative treatment have been reported by several authors although the methodology has varied.,,,, Okuyama, in a detailed study and follow up of 19 patients found that 47% had no pain and 58% had returned to hard work after 5 months. The follow up kyphotic angle was 12 degree and there were no serious complications. Clohisy et al demonstrated better neurological recovery by early anterior decompression. Pain in the back is a very significant pointer in patients with burst fractures. Advocates of conservative treatment feel that back pain is minimal on long term results and it does not disturb the working ability. Reduction of kyphosis was not absolutely essential, as back pain did not co-relate with the degree of kyphosis.,, Krompinger et al felt that kyphosis of 30o is unacceptable and that surgical intervention is essential. In Okuyama's series, the kyphotic angle ranged from 5 to 45 degrees (average 12 degrees) but there was very little back pain. The experience in our series has been similar and over a period of last 25 years, we have not been able to establish satisfactory correlation between kyphosis and back pain or worsening of neurological deficit. In our earlier cases, three patients treated with posterior hartshill ring stabilization and anterior decompression and fusion without implants had developed more than 30 degree kyphosis (30 degrees, 36 degrees and 40 degrees), when they were assessed at eight months. All the three patients had complained of renewed back pain and felt they had become shorter in height. But there was no worsening of neurological deficit and hence they were treated with bracing. Similar problems have not arisen in the present series. Knight et al believe that a kyphotic angle of 20 degree or more should be treated surgically.
In burst fractures, anterior and middle columns are involved. Short segment pedicle screws and VSP plates stabilization from behind is simple and helps in decompression of the cord. Kyphosis is corrected effectively. The posterior approach was also extremely important in seven cases of dural laceration. Anterior surgery achieves more complete and reliable decompression with interbody strut graft fusion along the lines of axial loading which is so very important in the biomechanics of the spinal functioning in this region. Anterior surgery has better advantage of canal clearance than posterior pedicle screws and plates system. For early rehabilitation program aimed at shorter hospitalization and with intention to return the patient to full time work, it is necessary to have stiff anterior construct with iliac crest bone grafting and implants. Edelker et al showed that two motion segment stabilization with pedicle screws and VSP plates along with anterior bone grafting addresses effectively the anterior and middle columns. Hamilton showed that short segment pedicle fixation has powerful inherent correction properties. He also believed in posterior longitudinal ligamentotaxis with pedicle screws resulting in reduction of fractured body and clearance in bony encroachment of the canal.
No strong conclusions should be drawn to say that the results of surgery are superior to non operative treatment. However, anterior surgery along with posterior pedicle screw stabilization does give patients rigid stabilization, good clearance of the canal with satisfactory decompression of the spinal cord and allows early rehabilitation with shorter hospital stay and early return to work. With these findings it is proposed that surgical treatment providing a rigid spine capable of early bony arthrodesis should be advocated in all cases of unstable burst fractures in the dorso lumbar region irrespective of the status of neurological damage.