Adjacent Vertebral Refracture Prediction Model Based on Imaging Data After Vertebroplasty for Osteoporotic Vertebral Compression Fracture

Shufeng Shen,Xinmao You,Yingqing Ren,Senqi Ye
DOI: https://doi.org/10.1016/j.wneu.2024.07.169
2024-07-27
Abstract:Objectives: To establish a predictive model to evaluate the risk of adjacent vertebral refracture (VRF) after percutaneous kyphoplasty (PKP) for osteoporotic vertebral compression fracture (OVCF) based on perioperative imaging data. Methods: This study was a retrospective cohort study which established a predictive model of VRF after PKP for OVCF. Patients who underwent PKP for OVCF in our hospital between January 2018 and December 2020 were enrolled and divided into a refracture group and normal group. Perioperative imaging data including preoperative bone mineral density (BMD), fatty infiltration (FI%) of paravertebral muscle, sagittal parameters of the spine and pelvis, and recovery rate of vertebral height were collected. The prediction model is obtained by multifactor logistic regression analysis. Results: A total of 242 patients were included, including 23 cases in the VRF group and 219 cases in the normal group. There were statistical differences in BMD, FI%, recovery rate of vertebral height, and sagittal imbalance between the 2 groups. Receiver operating characteristic curve analysis of continuous variables showed that BMD ≤-2.80, FI% ≥40%, and recovery rate of vertebral height ≥ 10% were the cutoff values. Logistic regression analysis showed that BMD ≤-2.80, FI% ≥40%, and sagittal imbalance were independent risk factors for VRF. The area under the curve according to the predicted probability was 0.85 (P < 0.05). After simplifying the model, the total point of the model was 7 points, with a cutoff value of 5 points. Conclusions: The prediction model obtained in this study can predict refracture after PKP for OVCF early and effectively. It has an excellent predictive effect which is suitable for clinicians.
What problem does this paper attempt to address?