A Preliminary Algorithm Using Spine Measurement Software to Predict Sagittal Alignment Following Pedicle Subtraction Osteotomy

STUDY DESIGN Retrospective case series. OBJECTIVE To evaluate if spine measurement software can simulate sagittal alignment following pedicle subtraction osteotomy (PSO). METHODS We retrospectively reviewed consecutive adult spinal deformity patients who underwent lumbar PSO. Sagittal measurements were performed on preoperative lateral, standing radiographs. Sagittal measurements after simulated PSO were compared to actual postoperative measurements. A regression equation was developed using cases 1-7 to determine the amount of manual rotation required of each film to match the simulated sagittal vertical axis (SVA) to the actual postoperative SVA. The equation was then applied to cases 8-13. RESULTS For all 13 cases, the spine software accurately simulated lumbar lordosis, pelvic incidence lumbar lordosis mismatch, and T1 pelvic angle, with no significant differences between actual and simulated measurements. The pelvic tilt (PT), sacral slope (SS), thoracolumbar alignment (TL), thoracic kyphosis (TK), T9 spino-pelvic inclination (T9SPi), T1 spino-pelvic inclination (T1SPi), and SVA were inaccurately simulated. The PT, SS, T9SPi, T1SPi, and SVA all change with manual rotation of the film, and by using the regression equation developed with cases 1-7, we were able to improve the accuracy and decrease the variability of the simulated PT, SS, T9SPi, T1SPi, and SVA for cases 8-13. CONCLUSIONS Dedicated spine measurement software can accurately simulate certain sagittal measurements, such as LL, PI-LL, and TPA, following PSO. A number of measurements, including PT, SS, TL, TK, T9SPi, T1SPi, and SVA were inaccurately simulated. Our preliminary algorithm improved the accuracy and decreased the variability of certain measurements, but requires future prospective studies for further validation.