Vertebral-Pelvic Angle Settling: Quantifying T4-L1-Hip Axis Changes Between Intraoperative and Postoperative...

Retrospective cohort.

To evaluate changes in vertebral-pelvic angles (VPAs) between intraoperative and 6-week postoperative radiographs after fusions for adult spinal deformity (ASD).

VPAs are measures of sagittal balance that are independent of posture. Their magnitude is expected to remain relatively fixed within spinal constructs between intraoperative and postoperative alignment. However, no studies have examined VPA fidelity between intraoperative prone and postoperative standing radiographs or analyzed potential factors influencing VPA shifts.

We analyzed data for patients receiving spinopelvic fusions for ASD from at least L2-pelvis. Pre- and 6-week postoperative VPAs were measured on standing radiographs. Intraoperative VPAs were measured on prone radiographs obtained after screw tightening. Primary VPAs were L1-pelvic angle (L1PA) and T4-pelvic angle (T4PA). Uni- and multivariable regressions were performed to assess variables that may be associated with intra-to-postoperative VPA changes. Analyses were stratified by whether constructs controlled L1PA only (L1C; UIV: L2-T6) or T4PA+L1PA (T4L1C; UIV: T5 and above).

216 patients were analyzed (mean age 64.9±10.7 y, 68% female). 80 patients were L1C and 136 patients were T4L1C. Regardless of UIV, both L1PA (p≤0.0001) and T4PA (P<0.0001) increased on postoperative radiographs. L1PA changed by 2.94±2.5° and 3.43±2.5° for L1C and T4L1C, respectively. T4PA changed by 4.62±2.8° and 4.23±2.9° for L1C and T4L1C, respectively. T4PA-L1PA concordance similarly worsened (p≤0.007 for LIC and T4L1C). Multivariable analyses identified preoperative C2PA, 3-column osteotomy, rod material, rod diameter, rod number, and lumbar interbody use as significantly associated with ΔL1PA or ΔT4PA.

L1PA, T4PA, and T4PA-L1PA mismatch increased with upright posture postoperatively, even when stratifying by fusion length. Several variables were potentially related to these changes. These data support the notion of "VPA settling" whereby the realigned spine and construct cannot counteract some amount of reversion towards preoperative alignment. Anticipating these changes could have implications for intraoperative decision-making and alignment goals.

III.