Recurrent fractures after lumbar vertebroplasty: Biomechanical risk factors and a predictive model

This is the authors' abstract. We don't add key points for this paper.

To analyze risk factors for recurrent fractures after lumbar vertebroplasty from a biomechanical perspective, establish a prediction model, and provide scientific evidence for clinical precise identification of high-risk patients and development of individualized prevention strategies. A retrospective study design was adopted to collect clinical data from 176 patients with osteoporotic vertebral compression fractures (OVCFs) who underwent lumbar vertebroplasty treatment at our hospital from January 2020 to December 2024. Patients were divided into recurrent fracture group (42 cases) and non-recurrent fracture group (134 cases) based on whether recurrent fractures occurred postoperatively. Patient demographic characteristics, osteoporosis-related indicators, surgical parameters, laboratory examinations, and imaging data were collected. Vertebral morphological changes, bone cement distribution characteristics, and sagittal balance parameters were evaluated through imaging methods. Univariate and multivariate logistic regression analyses were used to screen independent risk factors, establish prediction models, and construct risk scoring systems. A total of 176 patients were followed up for an average of (24.3±8.7) months, with 42 cases (23.9%) developing recurrent fractures, including 29 cases (69.0%) of adjacent vertebral fractures. Recurrent fractures mainly occurred within 1 year postoperatively (78.6%). Univariate analysis identified 16 potential risk factors, and multivariate logistic regression analysis ultimately determined 5 independent risk factors: age ≥75 years (OR=2.520, 95%CI: 1.124-5.649), lumbar BMD T-score ≤-3.0 (OR=3.175, 95%CI: 1.454-6.930), history of previous vertebral fractures (OR=2.221, 95%CI: 1.025-4.810), severe vertebral compression >40% (OR=2.380, 95%CI: 1.028-5.510), and uneven bone cement distribution grade III (OR=2.972, 95%CI: 1.097-8.049). The area under the ROC curve of the prediction model was 0.812 (95%CI: 0.745-0.879), with sensitivity of 76.2% and specificity of 79.1%. A risk scoring system (0-12 points) was established, with recurrent fracture rates of 67.4% in the high-risk group (8-12 points), 28.1% in the moderate-risk group (4-7 points), and 8.3% in the low-risk group (0-3 points) (P<0.001). Age ≥75 years, severe osteoporosis, history of previous vertebral fractures, severe vertebral compression, and uneven bone cement distribution are independent risk factors for recurrent fractures after lumbar vertebroplasty. The prediction model established based on these factors has good predictive efficacy, and the risk scoring system helps clinicians precisely identify high-risk patients, providing important references for developing individualized prevention strategies.