P48. Mechanisms of lumbar spine ‘flattening’ in adult spinal deformity: defining changes in shape that occur relative to a normative population

Abstract

<h3>BACKGROUND CONTEXT</h3> Loss of lumbar curvature is often expressed as an angular regional loss. Previous work comparing adult spinal deformity (ASD) lumbar alignment to their age-and-PI adjusted normative value demonstrated that, contrary to general belief, a large proportion of the curvature is lost proximally (L1-L4). This study is a follow-up looking not only at regional angles, but also at the spinal contour collectively. <h3>PURPOSE</h3> Investigate the difference of lumbar shape between ASD and age-and-PI adjusted normative values. <h3>STUDY DESIGN/SETTING</h3> Retrospective review of an ASD registry. <h3>PATIENT SAMPLE</h3> A total of 119 asymptomatic volunteers and 362 ASD patients. <h3>OUTCOME MEASURES</h3> Sagittal alignment of the lumbar and thoracolumbar spine measures using vertebra pelvic angle from L5 to T10. <h3>METHODS</h3> Before studying ASD patients, 119 asymptomatic volunteers with full-body, free-standing radiographs were used to identify age-and-PI models of each vertebra pelvic angle (VPA) from L5 to T10, a validated methodology characterizing the detailed shape of the spine. Our study cohort was a registry of surgical primary ASD patients without coronal malalignment (SRS-Schwab Type=N). The formulas developed in the asymptomatic population were applied to the ASD group to calculate an age-and-PI normative spine shape for each patient. Loss of lumbar lordosis was defined as the offset between age-and-PI normative value and pre-operative spino-pelvic alignment. Patients were stratified into four groups by the amount of lordosis lost, ranging from "no loss" to "30° loss." Paired t-tests were performed to compare actual and normative VPA shapes within each group. <h3>RESULTS</h3> Out of 1,495 patients enrolled in this registry, 453 were primary cases, and 653 were categorized as a Schwab type N, leading to a cohort of 362 patients (age=64.4±13, 57.1% F). Pre-operative alignment demonstrated a large variability with a mean PI-LL of 15°±21, distal LL=31°±15, and PI=55°±13. Compared to their age-and-PI normative values, ASD patients demonstrated a significant lordosis loss of 17°±19 with the following distribution: 14.1% no loss (mean: 0.1±2.3), 22.9% with 10° loss (mean: 9.9±2.9), 22.1% with 20° loss (mean: 20.0±2.8), and 29.3% with 30° loss (mean:33.8±6.0). Comparison of the VPAs (and therefore the shape) between each LL group and the normative shape demonstrated that the "no loss" patients had a lumbar spine slightly anterior to the normative shape from L4 to T10 (VPA difference of 2°). The shape of the "small deformity" group (10°) superimposed on the normative one from L5 to L2 (VPA with p>0.1) and became anterior at the L1 level. As the lordosis loss increased, the offset between ASD and normative shapes began to propagate to the distal levels and became significant extending caudally to the L3 level for the "20° loss" group and further down to L4 for the more severe group. <h3>CONCLUSIONS</h3> As the deformity progresses and the loss of lordosis increases, the difference between ASD shape and normative shape happens first proximally and then progresses incrementally caudally with increasing deformity. Understanding the spinal contour and the location of this loss, in addition to regional parameters, may be key to achieving a sustainable correction by identifying optimal and personalized post-operative shape. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.