We reviewed 9 cases of sacral tumors with presacral extension. These included 2 chordomas. 1 metastatic renal cell carcinoma, 2 schwannomas (1 malignant, 1 benign), 1 neurofibroma, 1 neurofibrosarcoma, 1 aneurysmal bone cyst, and an exceedingly rare meningioma. The sex of the patients was not significant. The age of the patients at diagnosis ranged from 13 to 68 years (mean, 47 years). Initial symptoms of low back and radiating leg pain were present in all but 1 patient. The duration of symptoms prior to diagnosis ranged from 1 month to 9 years (mean, 2.6 years). A delay in diagnosis of 2 years or more occurred in 6 of the 9 patients. Progressive perineal numbness and/or sphincter dysfunction were seen in 6 patients, and a palpable rectal mass was noted in 6 of 9 patients. The efficacy of various diagnostic tests is presented, as are the surgical options—needle biopsy and anterior and posterior approaches. Despite improved radiographic imaging techniques, these unusual tumors are often diagnosed at an advanced stage, and may masquerade as discogenic radiculopathy. Late diagnosis contributes to the difficulty of surgical extirpation. Anterior and posterior surgical approaches involving general, orthopedic, and urological surgeons may be required.

To define the pathophysiology of spinal cord dysfunction associated with spinal epidural abscess formation, we developed an experimental model. Spinal epidural abscesses were produced in rabbits by injecting Staphylococcus aureus into the posterior thoracolumbar epidural space under direct vision. Progressive neurological deficits were detected in 18 of 20 animals; severe paraparesis or paraplegia occurred in 75%, and sphincter dysfunction occurred in 55%. Clinical data, including the results of plain spine roentgenography, myelography, and biochemical and bacteriological examination of the cerebrospinal fluid, were recorded. Epidural abscesses with varying degrees of spinal cord compression were confirmed pathologically in 95% of the experimental group. Spinal cord white matter changes included vacuolization, loss of myelin, and axonal swelling. The gray matter of the spinal cords was relatively preserved. There was no microscopic evidence of thrombosis or vasculitis in the major blood vessels supplying the spinal cords. Histopathological changes detected in the spinal cords were more consistent with direct compression of neural tissue than with infarction. The progressive clinical course and the histopathological changes in the spinal cord after compression by abscess closely resembled those of experimental compression of the spinal cord by epidural neoplasm.

An experimental model of spinal epidural abscess was developed in rabbits by injecting Staphylococcus aureus into the posterior thoracolumbar epidural space. This model has been shown to reproduce the neurological, bacteriological, and radiological aspects of the human disease. In this study, the effect of the infectious epidural mass on the vasculature of the spinal cord in paraplegic rabbits was studied using microangiographic techniques. The normal vascular anatomy of the rabbit spinal cord was defined in control experiments. Vascular proliferation was demonstrated in the epidural space surrounding the abscesses. Anterior and paired posterior spinal arteries remained patent in paraplegic rabbits with mild or moderate spinal cord compression and in some cases of severe compression. In animals with severe compression, the anterior epidural venous plexus remained patent, but the dorsal spinal vein was occluded. Occlusion of perforating arteries occurred only with extreme spinal cord compression. These data indicate that the initial neurological deficit associated with experimental spinal epidural abscess is not due to vascular thrombosis.

No abstract