Multicenter retrospective review.
To calculate overall incidence of pedicle screw “plowing” in adolescent idiopathic scoliosis (AIS) patients who underwent posterior spinal fusion (PSF). To identify risk factors for pedicle screw plowing and associated postoperative outcomes, including loss of correction and revision rate.
Summary of Background Data.
Curve correction of AIS generates perpendicular stresses that can cause pedicle screws to lose alignment and “plow” through pedicles craniocaudally.
We reviewed records of 1057 patients who underwent PSF for AIS from 2002 to 2015. Preoperative and first postoperative erect radiographs were evaluated by two observers to determine (1) presence of plowing and (2) subsequent loss of correction (LOC). Plowing was defined as more than 25° sagittal angulation compared with pedicle axis or entry of the most dorsal part of the screw outside the pedicle projection. LOC was defined as postoperative change in focal angulation of an instrumented spinal level, when in consensus of both reviewers. Bivariate analyses were performed (alpha = 0.05).
Nineteen thousand five hundred sixty nine screws were assessed across our cohort of 1057 patients. Both observers agreed that 48 patients (4.5%) demonstrated plowing of more than or equal to one pedicle screw. For 72 screws (0.4%), both observers noted plowing, most commonly through the cranial cortex of the pedicle (65/72 screws) and at the lowest instrument vertebra (LIV) (17/72 screws). Factors associated with plowing included larger curves (P = 0.02); lower mean pedicle screw density (P = 0.0003); skeletal immaturity as measured by open triradiate cartilage (P = 0.04); and younger chronological age at time of surgery (P = 0.04). LOC occurred in 13 patients, most commonly at LIV (P < 0.0001). Revision rate for loss of screw fixation was higher in the plowing group (P = 0.003).
Pedicle screw plowing occurred in 4.5% of AIS patients, especially in those skeletally immature and with decreased implant density. Plowing commonly occurred in the cranial direction and was associated with LOC, particularly at the LIV.
Level of Evidence: 3