Risk factor analysis for predicting vertebral body re-collapse after posterior instrumented fusion in thoracolumbar burst fracture

Publication date: Available online 20 July 2017
Source:The Spine Journal
Author(s): Hae-Dong Jang, Jae Chul Lee, Jae-Wan Soh, Sung-Woo Choi, Hyeung-Kyu Cho, Byung-Joon Shin
Background ContextIn the posterior instrumented fusion surgery for thoracolumbar (T-L) burst fracture, early postoperative re-collapse of well-reduced vertebral body fracture could induces critical complications such as correction loss, post-traumatic kyphosis, metal failure, often lead to revision surgery. Furthermore, re-collapse is quite difficult to predict because of the variety of risk factors and no widely accepted accurate prediction systems existed. Although load sharing classification (LSC) has been known to help to decide the need for additional anterior column support, this radiographic scoring system has several critical limitations.Purpose1) To evaluate risk factors and predictors for postoperative re-collapse in T-L burst fractures. 2) Through the decision-making model, we aimed to predict re-collapse and prevent unnecessary additional anterior spinal surgery.Study DesignRetrospective comparative study.Patient SampleTwo-hundred and eight (104 men and 104 women) consecutive patients with thoracolumbar burst fracture who underwent posterior instrumented fusion were reviewed retrospectively. Burst fracture caused by high energy trauma (fall from a height and motor vehicle accident) with a minimum one-year follow-up were included. The average age at the time of surgery was 45.9 years (range, 15 to 79). With respect to the involved spinal level, 95 cases (45.6%) involved L1, 51 involved T12, 54 involved L2, and 8 involved T11. Mean fixation segments were 3.5 (range, 2 to 5). Pedicle screw instrumentation including fractured vertebra had been performed in 129 patients (62.3%).Outcome MeasuresClinical data using self-report measures (visual analogue scale score), radiographic measurements (plain radiograph, computed tomography, and magnetic resonance image), and functional measures using the Oswestry Disability Index were evaluated.MethodsBody height loss of fractured vertebra, body wedge angle and Cobb angle were measured in serial plain radiographs. We assigned patients to the re-collapse group if their body height loss was progressed greater than 20% at any follow-up time compared with immediate postoperative body height loss; we assigned the remaining patients to the well-maintained group. The χ² test and t-test of SPSS were used for comparison of differences between two groups and multiple logistic regression analysis for risk factor evaluation. Through the decision tree analysis of statistical package R, decision-making model was composed and a cut-off value of revealed risk factors and re-collapse rate of each subgroup were identified. Present study was supported by University College of medicine Research Fund (University to which authors belong). There was no external funding source for this study. The authors have no conflict of interest to declare.ResultsRe-collapse occurred in 31 of 208 patients (14.9%). In this group, age, the proportion of male gender, preoperative height loss, and preoperative wedge angle were significantly greater than the well-maintained group. Multivariable logistic regression analysis identified two independent risk factors: age (adjusted odds ratio 1.084, p = 0.002) and body height loss (adjusted odds ratio 1.065, p = 0.003). According to decision-making tree, age (>43 years) was the most discriminating variable, and preoperative body height loss (> 54%) was the second. In this model, the re-collapse rate was zero in the ages less than 43 years and among the remains, nearly 80% patients greater than 54 % of body height loss belonged to the re-collapse group.ConclusionsThe independent predictors of re-collapse after posterior instrumented fusion for T-L burst fracture were the age at operation (> 43 years old) and preoperative body height loss (> 54%). Careful assessment using our decision-making model could help to predict re-collapse and prevent unnecessary additional spinal surgery for anterior column support, especially in young patients.



http://ift.tt/2uGM1Xh