Posterior instrumentation and fusion of the thoracolumbar spine for treatment of neuromuscular scoliosis.

We reviewed the clinical and technical outcomes of 25 patients with neuromuscular scoliosis, who were treated by Luque instrumentation and posterior spinal fusion from the upper thoracic spine to L5 between 1981 and 1988. A mean curve correction of 46% was obtained operatively with a mean 8 degrees loss of correction during the follow-up period that ranged from 1.9 to 9.4 years (mean, 5.5). Pelvic obliquity was improved 50% from a mean of 16.1 degrees to a mean of 8.1 degrees in 24 patients for whom data were available. At final follow-up, the mean pelvic obliquity increased to 11.4 degrees with only two patients increasing > 8 degrees. The cause for major postoperative increase in pelvic obliquity was continued anterior spinal growth with torsion of the fusion mass and was not related to changes limited to the L5-S1 motion segment. Posterior fusion and instrumentation from the upper thoracic spine to L5 without anterior fusion provides adequate correction and control of spinal deformity for many patients with cerebral palsy. Those patients with significant growth remaining, or with severe deformities, may benefit by preliminary anterior release and fusion or inclusion of the pelvis and sacrum.

Induction and characterization of an interface tissue by implantation of methylmethacrylate cement into the posterior part of the cervical spine of the dog.

After the implantation of methylmethacrylate cement into the posterior part of the cervical spine of the dog, a thick layer of connective tissue forms at the bone-cement interface. The tissue is six to eight millimeters thick and in all animals it surrounds the dorsal and lateral aspects of the masses of implanted cement, grows between the undersurface of the cement and the bone of the posterior elements, and completely covers that bone. This tissue was examined by light and electron microscopy and its collagenous components were extracted and analyzed biochemically by gel electrophoresis. Specific extracellular matrix proteins in the tissue at the bone-cement interface were also localized by immunohistochemistry. The tissue at the host-cement interface contained zones of fibrocytes and plump and teardrop-shaped cells within a collagenous matrix. Type-I, Type-III, and Type-V collagen were extracted and were identified by gel electrophoresis. Type-V collagen and fibronectin were localized predominantly around the plump and teardrop-shaped cells. Type-IV collagen and laminin were localized predominantly in an area just beneath the teardrop-shaped cells at the surface of the tissue overlying the cement, suggesting that a basement-membrane-like tissue had formed in this area.

Cervical ligamentous instability in a canine in vivo model.

A canine in vivo model of midcervical ligamentous instability was developed by dividing the anterior longitudinal ligament, anulus fibrosus, and all posterior ligamentous structures including the ligamentum flavum. The natural history of healing in the model, the effect on its healing by an adjacent one-level arthrodesis, and the effect of a one-level arthrodesis on normal adjacent ligamentous structures were studied radiographically, mechanically, and histologically. The authors determined that healing takes place primarily by anterior scar formation in their instability model but not to a degree sufficient to recreate normal mechanical stability. After three months, healing in the model was not affected by an adjacent arthrodesis; however, acutely, instability apparently was increased as three animals became quadriplegic between the second and fourth postoperative days. Arthrodesis did not affect adjacent normal ligamentous structures, during this period. Incomplete healing in the authors' model supports those who advocate arthrodesis as the treatment of choice for destabilizing cervical ligamentous injury. The authors previously reported the case of a patient who sustained bilateral facet dislocations adjacent to an arthrodesed segment and questioned whether this resulted from a stress-concentrating effect. This study indicates that this could well have been the case acutely. Thus, inadvertent exclusion of an unstable segment from an arthrodesis has potentially catastrophic results. Finally, the authors also have previously questioned whether arthrodesis of a midcervical segment could lead to instability of adjacent normal segments. This project does not support such a concern, at least for the three postoperative months of study.