BIOMECHANICS OF SPINAL INJURIES PRODUCED BY VERTICAL COMPRESSION
Abstract
Fifty-six cadavers were studied to analyze failure mechanics of the spine using vertical compression loading. In addition, six other spines were studied with three- or four-point loading. The cervical spine was loaded to failure in 13 cadavers. Ten were studied as isolated columns; the other three were intact. Forces were directed axially, or spines were loaded in preflexion or extension. Loads required to produce injury ranged from 645 to 7439 N. Flexion injuries were produced at an average value of 1989 Newtons and extension at 2196 N; failures with axial loading averaged 5969 N. The column alterations produced, particularly the ligament avulsions seen, correlated only partially with the force direction. Observed alterations were clinically consistent, however. The thoracolumbar spine was studied in 29 cadavers. Eighteen columns and five intact cadavers were subjected to flexion compression loading and six columns to three- or four-point bending. The most common levels of fracture were T11 and T12. Isolated columns failed at 2360 N (556-5275 N), and the intact cadavers at 1737 N (range 1110-2750 N). Mean bending moments were 168 Nm for the isolated columns, and 187 Nm for the intact cadavers. Differences in loads and moments between isolated columns and intact cadavers were not statistically significant. The three spines subjected to three-point bending failed at an average of 1761 N; mean failure was 2716 N for four-point bending. Mean bending moments were 115 and 136 Nm, respectively, and were significantly different than those reported for column loading. The injuries provided by (vertical) compression were reproducible and similar to those routinely seen. In nine isolated thoracolumbar column and five intact cadavers, the flexion/compression technique was applied to compare the failure characteristics of spines instrumented sequentially with modified Weiss springs, Harrington distraction rods, and Luque rods. Mean initial spine failure was 1833 N; spines instrumented with the springs failed (with production of spinal angles equal to those seen after initial fracture) at 1128 N (54% of control) by allowing bending of the spine. Harrington rodded-spines failed at 859 N (42%), by hook extrusion and lamina fractures. Luque rods failed at 83% of control at acute angles and provided the most rigid stabilization, followed by the modified springs. The consistent nature of the trauma and clinical pertinence make compression loading of the spine a valuable technique in spine analysis.
This paper has been withdrawn.