Harrington instrumentation for thoracic and lumbar vertebral fractures

Surgical treatment of traumatic fractures of the thoracic and lumbar spine: a systematic review of the literature on techniques, complications, and outcome

STUDY DESIGN

A systematic review of the literature, pertaining surgical treatment of traumatic thoracic and lumbar spine fractures, was performed.

OBJECTIVES

To provide information on surgical techniques for traumatic spine fracture management, their respective performance and complication rates, based on previously published information.

SUMMARY OF BACKGROUND DATA

The treatment of traumatic fractures of the thoracic and lumbar spine remains controversial. There is insufficient evidence in the literature to choose between the various surgical options. In absence of conclusive studies, a systematic review can be an alternative to obtain more convincing information.

METHODS

Full-text papers from 1970 until 2001 were included if strict inclusion criteria were met. Five surgical subgroups were recognized: posterior short-segment (PS), posterior long-segment (PL), reports on both posterior short- and long-segment (PSL), anterior (A), and anterior combined with posterior (AP) techniques. Clearly defined and generally accepted parameters were scored and subsequently analyzed. The preoperative injury severity of the surgical groups was compared. The neurologic, radiologic, and functional outcome and complications of all groups were assessed.

RESULTS

A total of 132 papers, the majority being retrospective case-series, were included representing 5,748 patients. The preoperative injury severity revealed an inequality between the subgroups. Partial neurologic deficits had the potential to resolve irrespective of treatment choice. None of the five techniques used was able to maintain the corrected kyphosis angle. The functional outcome after surgery seems to be better than generally believed. Complications are relatively rare.

CONCLUSIONS

In general, surgical treatment of traumatic spine fractures is safe and effective. Surgical techniques can only be compared using randomized controlled trials.

Low lumbar burst fractures: comparison between conservative and surgical treatments

Twenty-two low lumbar burst fractures (L3-L5) were treated, with an average follow up of 56.2 and 39.0 months in the conservative and surgically treated groups, respectively. Twenty patients were available for review; seven were treated conservatively and 13 were stabilized surgically. All patients were evaluated clinically for work status, activity level, residual pain, and subsequent development of neurologic symptoms. Roentgenograms were reviewed for severity of initial fracture, canal compromise, and maintenance of initial correction. In general, neurologically intact patients in both groups returned to similar postinjury employment levels. Persistent back pain was found to be more disabling in the surgically treated group, in which a fusion incorporating four or five lumbar segments was performed. There was no evidence of significant loss of initial reduction, and no patients experienced late neurological compromise in the surgical group. An average follow-up kyphosis of 9.2 degrees and 31% loss of vertebral height were observed in the conservative group, while a follow-up lordosis of 1 degree and 19% loss of vertebral height were observed in the surgical group. Conservative treatment of low lumbar burst fracture is a viable option in neurologically intact patients, but loss of lordosis and vertebral height may persist. Biomechanical and anatomic characteristics of the low lumbar spine differ from the thoracolumbar region and may account for the inherent stability of these injuries. If surgery is chosen, a long fusion with distraction instrumentation should be avoided in the low lumbar spine. A short rigid fixation with pedicular instrumentation may be of greater benefit.

Differentiation of spinal damage through compression mechanism

In the years 1965 to 1986, 986 patients with spine fractures caused by compression were treated at the Rehabilitation Clinic. This communication presents an analysis of those patients, taking into account the significant differences between typical compression fractures and explosion fractures, the latter are, as a rule, complicated by considerable or complete spinal cord damage. The author has pointed to the differences in the indication for surgery, different prognosis and significantly different results of treatment. The author is of the opinion that, due to those differences, the two kinds of spinal fractures should be considered separately with regard to the analysis of the mechanism of injury, the prognosis and the assessment of the results of treatment.

Percutaneous interbody osteosynthesis in the treatment of thoracolumbar traumatic or tumoural lesions. A review of 51 cases

The authors describe a technique of percutaneous interbody osteosynthesis applicable to the dorsal and lumbar spine. 51 patients were so treated for different aetiologies: traumatic conditions (35 cases) and tumoural lesions (16 cases). The material used consisted of special instruments that are positioned in double obliquity by a percutaneous posterolateral approach. A posterior approach limited to the pathological focus was used jointly whenever a graft or a decompression was necessary (19 cases). Several types of anaesthesia were used (local, local-regional, general, neuroleptanalgesia). The patients benefited by the advantages that usually accompany percutaneous techniques. The advantages and limitations of the method are discussed.

Recurrence of deformity after removal of Harrington's fixation of spine fracture. Seventy-six cases followed for 2 years

The radiographic result was assessed in 76 patients with acute unstable fractures of the thoracic or lumbar spine admitted during the years 1977-1984, and who were managed by early reduction and stabilization using Harrington distraction rods and a three-segmental posterolateral fusion. The radiographs were analyzed for anterior and posterior heights plus sagittal and frontal widths of the fractured vertebral body and the angles of kyphosis and scoliosis of the spine. All the measurements were made at admission, immediately postoperatively, and at the latest follow-up at least 3 months after removal of the rods, which was done as a routine procedure 6-12 months after the accident. The mean follow-up was 29 months. The posterior height and sagittal width of the vertebral bodies were best restored, whereas the initially well-reduced anterior height and the angle of kyphosis often had returned to values close to those seen on admission. The best anatomic restoration was obtained in rotation-dislocation injuries of the thoracic and thoracolumbar spine, and was poorest in burst fractures of the lumbar spine.

Neurological outcome after surgery for thoracic and lumbar spine injuries

New techniques for internal fixation, employed by neurosurgeons and orthopaedic surgeons, provide reduction and stabilization with early mobilization of patients after injuries of the thoracolumbar spine. It has, however, appeared difficult to prove that the neurological outcome after surgery may be superior to conservative treatment. In this study the neurological results after surgery are evaluated in 30 patients with injuries of the spinal cord and/or cauda equina. Canal encroachment was found on tomograms or CT in all patients. The unstable fractures, dislocations or fracture-dislocations were all reduced and stabilized by the Harrington distraction device. However, peroperative inspection revealed that the Harrington instrumentation alone may be inadequate in achieving intraspinal decompression. Additional anterior decompression by posterolateral approach was necessary in 13 of the 30 patients (43%). In this series there was an unexpectedly high rate of dural tears. These were found in 8 patients (27%) and nerve roots (cauda equina) appeared to have herniated through the tear in 6 of the 8 patients. Significant improvement of the neurological deficit was obtained in 22 of the 30 patients (73%), and the improvement rate in patients with incomplete lesions of the spinal cord was 83% (25 of 30 patients). In this series there was a definite trend toward greater neurological recovery when compared to the results obtained in patients treated conservatively. The importance of a neurosurgical approach in diagnosis and treatment of these patients is stressed.

Transpedicular segmental fixation: description of a new procedure

The treatment of unstable fractures of thoracolumbar vertebrae is a controversial issue in the orthopedic community. The various methods employed for operative stabilization of these injuries have to date been found to have major disadvantages. A new therapeutic principle and a new device for stable internal fixation are presented.

Operative fixation of fractures of the thoracic and lumbar vertebrae by Williams plates with reference to late kyphosis

Thirty patients with fractures of the thoracic and lumbar spine, treated by the Williams plating technique (with or without fusion) between 1966 and 1978, were reviewed and followed up for an average of 6.5 years (range 2-14 years). The average degree of primary traumatic kyphosis was 11.4 degrees (range 0-33 degrees), and the average degree of late deformation kyphosis during follow-up was 17.4 degrees (range 0-30 degrees). The average loss of operative correction during follow-up was 9.8 degrees. Technical late complications, such as loosening of screws, occurred in half of the patients, but they did not affect the degree of spinal deformation. The corrective result, with reference to late deformity (kyphosis), largely correlated with the degree of primary traumatic kyphotic deformation and the severity of the neurologic lesion. The best late correction result was achieved with moderate primary kyphosis in patients with bursting fractures. Spinal fusion tended to give better results, especially in anterior wedge fractures, in maintenance of the correction of kyphosis, while laminectomy tended to give poorer results. None of the 15 patients with primary complete paraplegia showed improvement, while 80 per cent of patients with incomplete paraplegia made a moderate recovery.

Analysis of 75 operated thoracolumbar fractures and fracture dislocations with and without neurological deficit

Seventy-five surgically treated patients with thoracolumbar fractures and fracture dislocations, operated on between 1978 and 1982 at the Orthopedic Department of the University of Basel, were analyzed. The follow-up ranged from 18 months to 6 years. There were 45 men and 21 women, and 60% of the patients were not more than 30 years old. Additional injuries were common: 30% of the patients had craniocerebral injuries and 20% were polytraumatized. Ninety-six percent of all patients reached a hospital within 6 h, but only 23% initially presented at a center for spinal surgery. Sixteen patients had anterior surgery (fusion alone or with plating), and two of these had laminectomy as a second operation. Fifty-seven patients had posterior surgery, in 34 cases combined with a laminectomy. The Harrington instrumentation was used 45 times (29 distraction, 14 compression, and two combinations of distraction and compression rods). Luque rods with segmental sublaminar wiring was used seven times, the locking-hook distraction-rod system of Jacobs twice, and miscellaneous procedures five times. A total of 24 patients (greater than 30%) presenting neurological deficits improved postoperatively. None of the 18 patients with normal neurological findings deteriorated during the operation. Neurological improvement was seen more frequently after early than after delayed surgery, but the difference was not statistically significant. Laminectomy had no statistically significant effect on postoperative neurological status. Twenty-two patients required reoperation because of insufficient or failed instrumentation. Luque instrumentation had the highest rate of reoperations. Anterior surgery did not prove superior to posterior procedures. Hospitalization and immobilization time was significantly reduced with surgery for the neurologically normal or minimally damaged patients, but not for completely or incompletely paraplegic patients. Postoperative back pain occurred in 22 patients, of whom 14 had nonanatomic postoperative reductions. Complications directly due to the surgery were rare. It is our conclusion that the instrumentation used in this series was not good enough to be proposed for standardized use. These technically unsatisfactory results induced the development of the internal fixator system in the senior author's (E.M.) department.