Risk factors for rod fracture after posterior correction of adult spinal deformity with osteotomy: a retrospective case-series

Background: Osteotomies including pedicle subtraction (PSO) and/or Smith-Peterson (SPO) are used to facilitate surgical correction of adult spinal deformity (ASD), but are associated with complications including instrumentation failure and rod fracture (RF). The purpose of this study was to determine incidence and risk factors for RF, including a clinically significant subset (CSRF), after osteotomy for ASD. Methods: A retrospective review of clinical records was conducted on consecutive ASD patients treated with posterolateral instrumented fusion and osteotomy. Seventy-five patients (50 female; average age, 59) met strict inclusion/exclusion criteria and follow-up of >= 1 year. Data was extracted pertaining to the following variables: patient demographics; details of surgical intervention; instrumentation; and postoperative outcomes. Patients were divided into two subgroups: 1) rod fracture (RF) and 2) non-RF. The RF subgroup was further divided into CSRF and non-CSRF. Odds ratios (OR) were calculated to evaluate the association between risk factors and RF. The chi(2)-test was used to define P-values for categorical variables, and T-test was applied for continuous variables, P-values <= 0.05 were considered significant. Results: Incidence rates of RF were: for entire population, 9.3 % (95 % Cl: 2.7 %; 15.9 %); for PSO, 16.2 % (95 % Cl: 4.3; 28.1); and for SPO, 2.6 % (95 % Cl: 0 %; 7.7 %); the OR of PSO versus SPO was 7.2 (95 % Cl: 0.8; 62.7, P = 0.1). CSRF incidence was 5.3 % (95 % CI: 0.2 %; 10.4 %). Significant risk of RF was revealed for following factors: fusion construct crossing both thoracolumbar and lumbosacral junctions (OR = 9.1, P = 0.05), sagittal rod contour >60 degrees (OR = 10.0, P = 0.04); the presence of dominos and/or parallel connectors at date of rod fracture (OR = 10.0, P = 0.01); and pseudarthrosis at >= 1 year follow-up (OR = 28.9, P < 0.001). Statistically significant risk of CSRF was revealed for fusion to pelvis (P = 0.05) and pseudarthrosis at >= 1 year follow-up (OR = 50.3, CI: 4.2; 598.8, P < 0.01). Conclusions: The risk of RF after posterolateral instrumented correction of ASD with osteotomy had statistically significant association with the following factors: pseudarthrosis at = 1 year follow-up; sagittal rod contour >60 degrees; presence of dominos and/or parallel connectors at date of fracture; and fusion construct crossing both thoracolumbar and lumbosacral junctions. Statistically significant risk for the CSRF subset was fusion to the pelvis and pseudarthrosis at >= 1 year follow-up.