Factors influencing neurological recovery in burst thoracolumbar fractures.

Published online: Sep 27 1995

G K Dendrinos, J G Halikias, P N Krallis, and A Asimakopoulos.

First Orthopedic Department, Athens General Hospital, Greece.


The association between the thoracolumbar vertebrae fracture pattern, treatment and neurological recovery was estimated. Sixty-three patients with burst fractures at the T11 to L2 vertebral level and associated neurological deficit were evaluated by plain roentgenograms, CT scan and a quantitative neurological examination. The parameters used were percent canal compromise, location of the retropulsed middle column fragment, kyphosis, type of treatment, and neurological recovery. The follow-up varied from 24 to 84 months (mean 44 months). Treatment was conservative in 15 patients and surgical in 48 patients. Posterolateral decompression was carried out in 26 patients. The severity of the initial paralysis did not correlate with the initial fracture pattern except perhaps for Frankel A cases. Neurological recovery did correlate with the initial kyphosis but not with the amount of canal compromise or the location of the middle column fragment. Neurological recovery did not correlate with decompression. Improvement of paralysis was associated with restoration of the sagittal spine alignment. From the patients with greater than 5 degrees correction of kyphosis the majority improved neurologically. If the correction of the kyphosis was less than 5 degrees the recovery was poor regardless of the method used. We assume that the initial paralysis in burst fractures with severe kyphosis is partially caused by permanent cord or root damage and partially by neuroapraxia from angulation of the neural structures and their vessels. With reduction of the fracture and correction of the kyphotic deformity, spinal cord, roots and their vessels become lax, and the chances for neurological recovery increase significantly.