Magnitude, Location, and Factors Related to Regional and Global Sagittal Alignment Change in Long Adult Deformity Constructs: Report of 183 Patients With 2-Year Follow-up.

Abstract

This is a retrospective review of a prospective multicenter adult spinal deformity (ASD) database. To quantify the location and magnitude of sagittal alignment changes within instrumented and noninstrumented spinal segments and to investigate the factors associated with these changes after surgery for ASD. Spinal realignment is one of the major goals in ASD surgery and changes in the alignment are common following surgical correction. Inclusion criteria: operative patients with age above 18, coronal Cobb angle ≥20 degrees, sagittal vertical axis (SVA) ≥5 cm, pelvic tilt ≥25 degrees, and/or thoracic kyphosis ≥60 degrees. revision surgery 6 weeks postoperatively. Standard sagittal radiographic spinal deformity parameters were evaluated. Changes in sagittal parameters between 6 weeks and 2 years postoperatively were assessed within and outside instrumented segments. Associations between changes in sagittal alignment and age, preoperative SVA, rod diameters, rod material, presence of 3-column osteotomy, and the use of interbody fusions were evaluated. Patients were also stratified by >5- and >10-degree changes in alignment. In total, 183 patients (male:29, female:154, average age: 56±14.8 y) met inclusion criteria. A total of 45(24.6%) patients had increase in pelvic tilt >5 degrees, 74(40.4%) had increase in pelvic incidence and lumbar lordosis (LL) >5 degrees, and 76 (41.5%) had increase in SVA >2 cm. Mean change of thoracic sagittal alignment was +8 degrees; 70 (60%) patients had increases of >5 degrees and 31 (27%) had increases of >10 degrees. Noninstrumented thoracic segments had significantly more increase than instrumented thoracic segments (P=0.02). Mean loss of LL was -6 degrees; 49(47%) patients had worsening >5 degrees and 13(13%) >10 degrees. Noninstrumented lumbar segments had significantly less loss of lordosis than instrumented segments (P<0.01). Risks for loss of LL were: age 65 years and above [odds ratio (OR) 9.4; 95% confidence interval (CI), 3.5-25.2; P<0.01], preoperative SVA>5 cm (OR, 2.4; 95% CI, 1.3-4.4; P<0.01), and lumbar interbody fusion (OR, 2.3; 95% CI, 1.2-4.2; P<0.01). Smaller rods (4.5 mm) were associated with a lower probability of worsening LL compared with 5.5-mm rods (OR, 0.15; 95% CI, 0.04-0.58; P<0.01) and 6.0-mm rods (OR, 0.36; 95% CI, 0.18-0.72; P<0.01). The presence of a 3-column osteotomy and rod material were not significant factors in alignment changes (P>0.05). After correction of ASD, increases in thoracic and decreases in lumbar alignment is common. Loss of thoracic sagittal alignment primarily occurs in noninstrumented thoracic segments, whereas instrumented lumbar levels in elderly patients ( above 65 y) with high preoperative SVA, interbody fusions, and larger rods have significantly higher rates of postoperative sagittal alignment changes in the lumbar spine.