A biomechanical study shows the direction of compression influences the amount of lordosis gained in lumbar fusion.

Study Design: Biomechanical model study.

Background: Lumbar lordosis is usually lost in the degenerative process, and when lumbar fusion is required, its restoration is one of the modern metrics of a successful operation. We sought to investigate the hypothesis that changing direction of compression during surgical fusion, would gain more lordosis.

Methods: Using a biomechanical Sawbones™ model we inserted polyaxial pedicle screws from S1 to L4. A rod was placed in the screws without requiring reduction. Markers were attached to the spinous processes to allow photographic analysis of lordosis. Two methods were compared. Method A - caudal screws were locked first and compression proceeded in a cranial direction prior to locking. Method B - cranial screws were locked first and compression proceeded caudally. Increasing levels of surgical invasiveness were tested; intact, interbody cage, inferior facet resection, and Ponte resection and using different rods including: lordotic, hyperlordotic and straight.

Findings: Method B demonstrated to be consistently superior to Method A, regardless of the type of rod used and for every level of surgical invasiveness performed. (P < 0.001).

Interpretation: locking the top screws first was a consistently superior method of compression, gaining more lordosis. To explain this finding we suggest the following: During posterior compression of pedicle screws along a fixed rod, screw motion is limited by the conflict between the fixed lordotic rod position, and the need for the moving screw to move in a kyphotic arc which is determined by the cage which acts as a pivot point.