Sacral fractures: an important problem. Retrospective analysis of 236 cases

Abstract

Sacral fractures, often undiagnosed and untreated, frequently result in neurologic symptoms and deficits to the lower extremities and urinary, rectal, and sexual dysfunctions. These same neurologic problems often remain the major chronic sequelae after the more obvious pelvic trauma lesion has healed. Specific treatments aimed at neurologic problems are available and may allow the patient functional recovery. This is illustrated by anatomic observations on the sacrum in 39 cadavers showing the relationship among sacral nerve roots within their foramina. These observations were valuable for a retrospective study of 236 consecutive patients with sacral fractures in a series of 776 patients with pelvic injuries. A new classification of sacral fractures evolved from this study and provided a better understanding of the mechanisms responsible for the associated neurologic symptoms. The classification is based on the direction, location, and level of sacral fractures. Three different zones were identified as having characteristic clinical presentations: Zone I, the region of the ala, was occasionally associated with partial damage to the fifth lumbar root. Zone II, the region of the sacral foramina, is frequently associated with sciatica but rarely with bladder dysfunction. Zone III, the region of the central sacral canal, is frequently associated with saddle anesthesia and loss of sphincter function. Routine pelvic roentgenograms were almost useless in identifying the pathologic process in sacral injuries with neurologic symptoms. Ferguson views, tomograms, and particularly computed tomography scans were crucial for understanding these injuries. Cystometrography was most helpful in positively identifying fractures causing neurogenic bladders. Cystometrograms should be ordered routinely in Zone III injuries. Preliminary observations suggest that surgical decompression permitted significantly better neurologic recovery than nonsurgical methods.