Alexander T Yahanda, Tim T Bui, Karan Joseph, Samuel Vogl, Munish C Gupta, Jeffrey Hills, Eric O Klineberg, Gregory M Mundis, Christopher P Ames, Han Jo Kim, Michael P Kelly, Camilo A Molina
LLA and SS were correlated pre-and postoperatively for all patients. Correlation strength and predictive ability was superior for patients with aligned T4-L1-hip axes, highlighting the importance of proper alignment of the arches cranial to the LLA. Focusing on the LLA and T4-L1-hip axis should be a priority for surgical planning to optimize prediction of SS (and thus elimination of pathologic pelvic retroversion) after LLA modification regardless of fusion length or Roussouly curve type.
STUDY DESIGN: Retrospective cohort.
OBJECTIVE: To assess the combined effect of lower lordotic arch (LLA) and thoracolumbar (T4-L1-hip axis) alignment correction on the ability to predict sacral slope (SS) (i.e. pelvic retroversion) in adult spinal deformity (ASD).
SUMMARY OF BACKGROUND DATA: The Roussouly framework characterizes the spine through a series of lordotic and kyphotic arches. The LLA should be equivalent to SS, meaning that pelvic positioning should be predicted by lumbar lordosis modifications between the lordotic apex and S1. However, positive T4-L1-hip axis mismatches will drive persistent pelvic retroversion. We hypothesized that postoperative SS prediction from LLA correction will be highest among patients with aligned T4-L1 hip axes.
METHODS: 339 ASD patients receiving fusions of ≥L2-pelvis had spinopelvic parameters measured on preoperative and 6-week scoliosis radiographs to determine LLA-SS correlations. Uni- and multivariable analyses were used to ascertain the effect of the LLA, T4-L1-hip axis, and other factors on pre-/postoperative SS.
RESULTS: LLA and SS were correlated among all patients preoperatively (ρ=0.68, R2=0.47, P<0.0001) and postoperatively (ρ=0.74, R2=0.55, P<0.0001). Postoperative strength and predictive ability of these correlations improved when the T4-L1-hip axis was respected (ρ=0.82, R2=0.68, P<0.0001) versus non-respected (ρ=0.68, R2=0.47, P<0.0001). Multivariable analysis demonstrated that preoperative LLA (P<0.0001), T4-L1-hip axis alignment (P=0.02), pelvic incidence (P<0.0001) and sex (P=0.01) were predictors of preoperative SS. Postoperative LLA (P<0.0001), aligned T4-L1-hip axis (P=0.04), pelvic incidence (P=0.0005), and revision surgery (P=0.01) were significant predictors of postoperative SS.
CONCLUSIONS: LLA and SS were correlated pre-and postoperatively for all patients. Correlation strength and predictive ability was superior for patients with aligned T4-L1-hip axes, highlighting the importance of proper alignment of the arches cranial to the LLA. Focusing on the LLA and T4-L1-hip axis should be a priority for surgical planning to optimize prediction of SS (and thus elimination of pathologic pelvic retroversion) after LLA modification regardless of fusion length or Roussouly curve type.