Indications for Anterior Column Reconstruction in Treatment of Unstable Thoraco-lumbar Spinal Injuries

Introduction: A purpose of this study was to determine indications for anterior column reconstruction in treatment of thoraco-lumbar spinal injuries.

Methods: Forty-five patients who underwent posterior reconstruction for thoracic (Th1-Th10: 12 patients) or thoracolumbar/lumbar (Th11-L3: 33 patients) spinal injuries were reviewed with minimum two-year follow-up. Injury types were burst fractures in 16 patients, fracture-dislocations in 28, and seat-belt type injury in 1. Severe vertebral body comminution was associated with translational injury in 12 patients with fracture-dislocations. Surgical procedures of posterior reconstruction ranged from short (one above-one below) to long fusion using simple hook/rod or pedicle screw instrumentation with or without supplemental hook placement (claw hook technique). As one of the most important goals of spinal reconstruction for thoraco-lumbar spine injuries was restore and maintenance of physiological spinal alignment, radiological failure of the posterior reconstruction was defined as 10 or more degrees' correction loss of kyphosis. Following possible risk factors for failure of posterior reconstruction were assessed by multivariate logistic regression analysis: age, gender, degree of vertebral body collapsing in both sagittal and axial plane, magnitude of kyphosis correction, levels of injuries, extent of fusion area, and types of fixation devices. Degree of vertebral body collapsing was evaluated by scoring system proposed by McCormack, et al* in which "comminution score" in sagittal plane and "apposition score" in axial plane was independently graded from 1 to 3 points in order of severity.

Results: Significant correction loss (10 or more degrees) of kyphotic deformity occurred in 22 (48%) of 45 patients. In these 22 patients, an average correction loss was 14 degrees ranging from 10 to 27 degrees. Following important risk factors for failure of posterior reconstruction were detected by multivariate logistic regression analysis: (1) degree of vertebral body comminution in axial plane (Odds ratio per 1 point increase in apposition score: 16.9, p=0.0023) and (2) levels of injuries (Odds ratio of thoracolumbar/lumbar injuries: 152.2, p=0.0107). Other factors such as extent of fusion area or types of fixation devices were proved to be insignificant. Predicted probabilities of posterior reconstruction failure for thoracolumbar/lumbar spinal injuries with 2 or 3 points in apposition score were calculated to be 0.80 to 0.98.

Discussion: In the thoracolumbar and lumbar spine, anterior column reconstruction was mandatory for not only burst fractures but fracture-dislocations with moderate to severe vertebral body collapse. Pedicle screw fixation with or without claw hooks placement or extending fusion area was not effective solution to avoid failure following posterior reconstruction for thoracolumbar spinal injuries with anterior column insufficiency.

Reference: *McCormack T, Karaikovic E, Gaines RW: The load sharing classification of spine fractures. Spine (19): 1741-4, 1994.