431 The Part is not the Same as the Whole: Evaluating L4-S1 Versus Total Lumbar Lordosis Attainment for Adult Scoliosis Patients Fused With Pre-Bent Rods

INTRODUCTION: Recent studies suggest L4-S1 lordosis is similar across individuals whereas lumbar lordosis (LL) differs by sagittal shape classification. Pre-bent rods account for both parameters. METHODS: Patients >18 years old with idiopathic or degenerative scoliosis receiving posterior instrumented fusion from a single surgeon/single institution were reviewed. Those fused >6 levels with LIV at the sacrum/pelvis, lumbar PCOs, and PSRs were included. Revision cases were excluded. Demographics, operative/inpatient variables, pre and post-radiographic measurements were recorded. Post-operative sagittal alignment was compared to pre-op measurements and planned targets. Lumbar lordosis (LL) and L4-S1 restoration were evaluated in multivariable regression models that included scoliosis type and pre-operative spinopelvic parameters significantly different from planned. RESULTS: 29 patients were included. The average age was 66.1 ± 9.5 years; 21 were female (21/29, 72.4%). Average BMI was 26.8 ± 5.3 kg/m 2 . Degenerative scoliosis was more common than idiopathic (22/29, 75.9%). Pre-op standing LL and L4-S1 lordosis were –36.2 ± 10.8 and –41.9 ± 16.0, respectively. The fractional curve was 16.0 ± 12.3°. 9.3 ± 2.9 levels were fused per patient. Post-operative LL but not L4-S1 lordosis was significantly different from planned parameters (–5.1, 95% CI –[9.0–1.2], p = 0.01; –2.2, 95% CI –[5.4–9.8], p = 0.5). On multivariable linear regression, neither scoliosis type, magnitude of planned lumbar correction, pre-operative fractional curve or starting spinopelvic parameters correlated with magnitude of planned versus obtained LL (r 2 = .1, p = 0.59). L4-S1 LL was significantly correlated with amount of correction required (p < 0.005), but no other parameter (r 2 = .78, p < 0.005). CONCLUSIONS: Post-operative L4-S1 lordosis better matched planned goals versus overall LL and was significantly correlated with required correction to obtain target parameters. Scoliosis type, fractional curve, and other spinopelvic parameters were not significant.