Nonsurgical treatment of three-column thoracolumbar junction burst fractures without neurologic deficit.

STUDY DESIGN Retrospective review of the outcome of neurologically intact patients with three column thoracolumbar junction burst fractures that were treated nonsurgically. OBJECTIVE To further define the parameters for nonsurgical management of thoracolumbar junction burst fractures. SUMMARY OF BACKGROUND DATA Many texts list involvement of the posterior column as an indication for surgery and state that casting or bracing is mandatory. This has not been the authors' experience. METHODS Thirty-eight patients with nonpathologic, single-level burst fractures of T11, T12, L1, or L2, and with posterior element fractures were studied retrospectively. The selection criteria required that the patient be neurologically intact, that the pedicles and facet joints not be fractured or dislocated, and that the angle of kyphosis be less than 35 degrees. The extent of retropulsion, loss of vertebral height, and presence of lamina or process fractures were not criteria. No attempt was made to reduce the fracture. Patients were allowed immediate ambulation as tolerated. Jewett braces were used in nine patients, but no bracing was used in the remainder. RESULTS There were 22 males, 16 females. Median age 37 years (range, 16-65). Fracture involved both endplates in 16 patients (12 crush-cleavage type), the superior end-plate in 21, and the inferior endplate in 1. The hospital stay was from 2 to 18 days (median, 8 days). Follow-up averaged 4.1 years (range, 2.1-6.3). All patients remained neurologically intact. Eight patients had no pain, 24 had minimal pain, 4 had moderate pain, and 2 had moderate to severe pain. Twenty-nine of 38 patients (76%) were able to work at the same level. The initial kyphosis angle averaged 20 degrees (range 10-35 degrees). At follow-up it averaged 24 degrees (range 12-38 degrees). The maximum increase was 6 degrees. Some degree of retropulsed fragment resorption was noted in 35 cases. Complications were limited to transient urinary retention. There were no thromboembolisms, decubitus ulcers, or pulmonary complications. CONCLUSION Despite the use of less restrictive criteria, no brace, and early activity as tolerated, the results are similar to those obtained with more restrictive protocols. The presence of vertical lamina fracture, spinous process fracture, and transverse process fracture are not contra--indications. Activity restriction and bracing may be important for pain control but probably does not change the long-term result.