Comparison of the failure biomechanics of spinal fixation devices.

The failure biomechanics of Harrington distraction rods, modified Weiss springs, and Luque rods were studied in intact cadavers and isolated spinal columns using flexion-compression loading. Most spines fractured at T-11 or T-12 at applied loads ranging between 556 and 4220 newtons (mean = 1833 N). After Harrington distraction rod placement, the same spines failed at a mean load of 859 N (42% of control), always as a result of hook extrusion and often including lamina fracture (seven cases). When modified Weiss springs were used, the spines failed at a mean load of 1128 N (54% of control) by allowing the spine to bend to the initial failure angle; in most instances, deformities resolved when the load was reduced. Luque rods were tested in four specimens; these provided the most rigid stabilization and failed at 83% of control values. Modified Weiss springs often maintain spinal stability better than Harrington distraction rods.