Biomechanical effects of lateral pedicle wall violation and subsequent screw redirection on pedicle screw anchorage

PURPOSE: To investigate the biomechanical effects of lateral pedicle wall violation (LPWV) and subsequent screw redirection on pedicle screw anchorage under cyclic loading. METHODS: Twenty fresh-frozen human thoracolumbar vertebrae (T10-L1) were randomly allocated to two experimental groups (n = 10 each). In Group 1, each vertebra received one standard intrapedicular screw and one LPWV screw, with screw trajectories randomly assigned to the left or right pedicle. In Group 2, both pedicles were initially instrumented with LPWV screws; one screw per vertebra was randomly selected for redirection into a standard intrapedicular trajectory, while the contralateral LPWV screw was left in place. All screws were subjected to cyclic cranio-caudal loading with stepwise increasing compressive load until loosening. RESULTS: LPWV screws withstood significantly fewer load cycles until loosening than standard intrapedicular screws (5230 ± 2481 vs. 8205 ± 3027 cycles; p = 0.003), corresponding to a 36% reduction in anchorage. Redirected screws withstood significantly fewer load cycles until loosening than LPWV screws left in place (3686 ± 2105 vs. 5489 ± 2731 cycles; p = 0.018), representing an additional 33% reduction. CONCLUSION: LPWV substantially compromises pedicle screw anchorage under cyclic loading, even when the screw remains at least half contained within the pedicle and fully contained within the vertebral body. Redirection of LPWV screws using the same diameter does not restore fixation strength but is associated with a further reduction in anchorage, challenging the assumption that redirection inherently improves screw stability. Redirection of LPWV screws using screws of the same dimensions and design may therefore be considered with caution, particularly in the absence of neural or vascular compromise.