Pseudarthrosis in Adult Idiopathic Scoliosis Primary Fusions

Purpose: The purposes of this study were to analyze characteristics of adult idiopathic scoliosis pseudarthrosis sites according to the anatomic sites, fusion length, crosslinks and domino sites, global imbalance, curve size and segmental kyphosis, aging, smoking habits, comorbidity and SRS scores.

Methods: Primary instrumented fusion of adult idiopathic scoliosis (IS) treated by two surgeons between 1985 and 2000 at a single institution was evaluated. A retrospective chart and radiographic review of 111 patients (average age; 42.7 year, range 18.2- 62.9 year) with adult IS undergoing virgin instrumentation and fusion with a minimum 2-year follow-up (average 5.6 years; 2-12.8 years) was analyzed. No revision cases were included in this study.

Results: 16 patients had pseudarthroses (14%, 1 evident/15 revised, average age 46.9). 54% occurring between T9 and L1 and 81% (13 patients) presented with multiple levels involved (2-6). The number of fused vertebra was related to pseudarthrosis (11/56 patients more than 12 vertebral segments fused vs 5/55 12 vertebral segments or less fused). Nonunion cases according to the lowest instrumented vertebra were L1 (3 among 13), L3 (1/5), L4 (4/24), L5 (2/29), and S1 (6/27). Site of crosslinks or dominoes correlated more with a nonunion site in 11 patients (69%) and 12 levels (29%). Preoperative sagittal and coronal global imbalance did not increase the incidence of nonunion. A higher preoperative Cobb angle (>70º) and thoracolumbar kyphosis demonstrated a higher nonunion rate (7/16). Pseudarthroses were detected radiologically at an average 32.4 months (12-67 months, 7 >2years) and revised at an average 45.3 months (17-80 months) postoperatively. The most common radiologic finding in pseudarthroses was rod breakage (62.5%), progression of deformity (50%), disc space collapse (19%), hook pull-off (12.5%) and a halo sign around the screw (12.5%). Patient age at operation correlated with nonunion (6/17 patients (>55 years), 7/70 (35-55), 3/36 (18-35). Smoking history did not increase the incidence of nonunion (3 pseudos in 22 smokers and 13 in 85 nonsmokers). Pseudarthrosis was related with comorbidity (8/28 patients). Patients with pseudarthrosis had lower SRS 24 scores (average 84.3) than those without (average 91.9) (p<0 .05).

Conclusion: The thoracolumbar junction area demonstrated a high nonunion rate potentially because of the anatomical and mechanical transition. The number of fused vertebral segments rather than the lower instrumented vertebra was correlated more with pseudarthrosis. Mechanical factors such as larger coronal Cobb and thoracolumbar kyphosis were correlated with pseudarthrosis. Crosslinks, although used for mechanical strength of the implant construct, was often related with the nonunion site. Patients with pseudarthrosis had significantly lower SRS scores than those without.