Objective: Functional outcomes of neurologically intact patients with burst fractures may be dependent on final kyphosis at the end of treatment. Conservative treatment is indicated if an acceptable sagittal alignment of the spine can be anticipated. Thoracolumbar burst fractures are often grouped as a single entity where, in fact, anatomically distinct areas of the spine may behave differently owing to different biomechanical factors. The goal of this work was to evaluate differential behavior in terms of final kyphosis in anatomically distinct regions of the spine following stable burst fractures.
Methods: Prospective analysis of kyphosis in 60 patients treated conservatively for traumatic thoracolumbar burst fracture was conducted. Initial trauma supine radiographs were measured for initial kyphosis (Ki). Final kyphosis (Kf) in the upright patient was measured at the end of treatment. The Ki and Kf were plotted on a scatter graph; with use of linear regression analysis, a mathematical model was created to define a relationship between Ki and Kf based on anatomic level of the spine.
Results: The thoracolumbar spine behaved in two independent patterns with respect to Kf. Kf at the thoracolumbar junction (T11-L1) had a collapse pattern that could be approximated most accurately with the equation Kf = Ki + 0.5 Ki. At the midlumbar spine, L2-L3 level, a best-fit model for collapse was Kf = Ki + 4 degrees .
Conclusion: In this cohort of patients, fractures that were categorized as "stable" and not requiring surgery were studied for the purpose of determining differential collapse patterns in anatomically distinct areas of the lumbar spine. We have demonstrated that the thoracolumbar junction and the midlumbar spine behave differently biomechanically and recommend that these two anatomic levels be studied independently for research purposes.