Pelvic incidence on prone and supine pre-instrumentation films best predicts postoperative pelvic incidence in patients undergoing thoracolumbar fusion for deformity
Resident Mayo Clinic Mayo Clinic Rochester, MN, US
Introduction: Pelvic incidence (PI) is a key determinant of overall sagittal alignment. While conventionally held to be fixed parameter, increasing evidence suggests it to be variable. The present study sought to identify predictors of positional change in PI and to assess the association of preoperative radiographic characteristics and intraoperative maneuvers with postoperative PI.
Methods: Patients undergoing thoracolumbosacral fusion were identified and data were gathered on preoperative spinopelvic parameters, patient demographics, intraoperative surgical maneuvers, and final construct details. Preoperative spinopelvic parameters were measured on upright, supine, and prone films. Univariable analyses were performed to identify predictors of positional changes in PI and multivariable analysis was performed to identify independent predictors of 6-week postoperative PI.
Results: 138 patients were identified (mean age 66.0±8.7yr; 38.4% male). Pairwise comparisons showed significant differences in PI between preoperative standing and supine films (mean -3.2±4.2°; p< 0.001), preoperative supine and prone (3.5±4.7°; p< 0.001), preoperative and 6-week postoperative standing films (1.5±6.4°; p=0.01), and immediate and 6-week upright films (1.1±3.9°; p=0.007). Univariable comparisons showed PI decrease from standing to supine was predicted only by weight (87.0±15.0 vs 81.1±19.3kg; p=0.04); there were no significant predictors of increase in PI from standing to prone. Increase in PI from pre-op to 6-week standing films were predicted by lower preoperative PI class (p < 0.001), L5/S1 interbody placement (74.4 vs 52.0%; p=0.02),and change in PI from preop standing to supine (12.8 vs 39.8%; p=0.002) and from standing to prone (51.3 vs 13.3%; p< 0.001). Multivariable analysis showed 6-week postoperative PI was independently predicted only by preoperative supine PI (B=0.293; 95% CI [0.10, 0.48]; p=0.003) and prone PI (B=0.647; [0.44, 0.85]; p< 0.001). The final model predicted 77.8% of the variance in 6-week postoperative PI.
Conclusion : There are significant position-dependent changes in PI among patients with unfused pelvises and significant changes in PI are seen even after pelvic fixation with S2-alar-iliac screws. Pelvic incidence following thoracolumbar fusion was best predicted by PI on supine and prone films prior to instrumentation rather than upright films. Preoperative prone and supine films may therefore provide surgeons with the best information regarding the postoperative PI target that should be matched with intraoperative maneuvers.