Combination of the AO-Magerl and load-sharing classifications for the management of thoracolumbar burst fractures.

Abstract

The AO-Magerl classification is widely accepted for the appropriate management of patients with thoracolumbar burst fractures; however, it fails to assess the ability of the injured spine to withstand compressive loading and cannot predict instrumentation failure after short-segment posterior fixation. The load-sharing classification depends on the degree of comminution and apposition of bony fragments.We retrospectively classified according to both classifications 100 consecutive patients with 1-level thoracolumbar burst fractures treated nonoperatively or operatively within a 7-year period. Sixty neurologically intact patients (60%) were treated nonoperatively, 15 (15%) had short posterior instrumentation, 15 (15%) had short anterior instrumentation, and 10 (10%) had combined short posterior instrumentation and anterior strut grafting. Twenty-five of the 40 (60%) surgically treated patients had neurological impairment on admission. Clinical outcome was assessed using a pain and working ability scale. Mean follow-up was 52 months (range, 24-70 months). Function was satisfactory in 55 (92%) nonoperatively treated patients and in 33 (83%) surgically treated patients. Neurological improvement by American Spinal Injury Association (ASIA) grade was observed in patients with incomplete paraplegia (70% of neurologically impaired patients) who were treated operatively.The combination of AO-Magerl and load-sharing classifications provides for accurate selection of treatment, surgical approach, and length of instrumentation, and can guide the decision for additional anterior surgery.