Fractures of the thoraco-lumbar spine.

A personal prospective study of 98 consecutive patients presenting with neurological impairment and fractures or dislocations between the 9th thoracic and 2nd lumbar vertebrae bodies. Fifty-three patients underwent Harrington instrumentation, and 45 patients were treated recumbently. Neurological improvement was much better following Harrington rods in the complete paraplegia group but there was no difference in neurological recovery between the two groups in those with incomplete paraplegia. Forty-two patients who had been stabilised with Harrington rods underwent post-operative myelography or tomography to assess the adequacy of spinal decompression. The best results were in patients with adequate neural canal decompression. In 21 cases decompression had not been adequate, usually due to a stereotyped pattern in which the postero-superior aspect of the fractured body remained in the neural canal. All 21 underwent anterior decompression at an average of five months post injury. All the incomplete anterior decompression at an average of five months post injury. All the incomplete paraplegics (nine patients) regained the ability to walk, three of the 12 complete paraplegics improved and regained the ability to walk with bilateral ankle-foot orthoses. Neurological improvement was dependent upon the adequacy of spinal cord decompression and not upon Harrington rods. per se. Harrington rods alone were not adequate to decompress the spinal canal in 50 per cent of cases. The best results after anterior decompression occurred where neural compression was caused by a minimally displaced wedge fracture distal to T12.