The biomechanical consequences of rod reduction on pedicle screws: should it be avoided?

Background context: Rod contouring is frequently required to allow for appropriate alignment of pedicle screw-rod constructs. When residual mismatch is still present, a rod persuasion device is often used to achieve further rod reduction. Despite its popularity and widespread use, the biomechanical consequences of this technique have not been evaluated.

Purpose: To evaluate the biomechanical fixation strength of pedicle screws after attempted reduction of a rod-pedicle screw mismatch using a rod persuasion device.

Methods: Fifteen 3-level, human cadaveric thoracic specimens were prepared and scanned for bone mineral density. Osteoporotic (n=6) and normal (n=9) specimens were instrumented with 5.0-mm-diameter pedicle screws; for each pair of comparison level tested, the bilateral screws were equal in length, and the screw length was determined by the thoracic level and size of the vertebra (35 to 45 mm). Titanium 5.5-mm rods were contoured and secured to the pedicle screws at the proximal and distal levels. For the middle segment, the rod on the right side was intentionally contoured to create a 5-mm residual gap between the inner bushing of the pedicle screw and the rod. A rod persuasion device was then used to engage the setscrew. The left side served as a control with perfect screw/rod alignment. After 30 minutes, constructs were disassembled and vertebrae individually potted. The implants were pulled in-line with the screw axis with peak pullout strength (POS) measured in Newton (N). For the proximal and distal segments, pedicle screws on the right side were taken out and reinserted through the same trajectory to simulate screw depth adjustment as an alternative to rod reduction.

Results: Pedicle screws reduced to the rod generated a 48% lower mean POS (495±379 N) relative to the controls (954±237 N) (p<.05) and significantly decreased work energy to failure (p<.05). Nearly half (n=7) of the pedicle screws had failed during the reduction attempt with visible pullout of the screw. After reduction, decreased POS was observed in both normal (p<.05) and osteoporotic (p<.05) bone. Back out and reinsertion of the screw resulted in no significant difference in mean POS, stiffness, and work energy to failure (p>.05).

Conclusions: In circumstances where a rod is not fully seated within the pedicle screw, the use of a rod persuasion device decreases the overall POS and work energy to failure of the screw or results in outright failure. Further rod contouring or correction of pedicle screw depth of insertion may be warranted to allow for appropriate alignment of the longitudinal rods.