Scholarlink

Purpose <p>Patients undergoing percutaneous vertebroplasty (PVP) are at a significant risk of developing adjacent vertebral fractures (AVF) following osteoporotic vertebral fractures (OVFs). Therefore, establishing a risk scoring system for AVF is crucial for timely intervention.</p> Materials and methods <p>Patients who underwent PVP for OVFs at our hospital from June 2016 to August 2022 were retrospectively included, with cases from June 2016 to May 2021 used for model training and June 2021 to August 2022 for validation. Patients were divided into AVF and non-refracture (NRF) groups according to whether AVF occurred after treated within 24 months in both cohorts. Potential predictors were compared, and independent risk factors were identified using multivariate logistic and LASSO regression. A risk scoring system was developed and evaluated for discrimination (ROC curve), calibration, and clinical utility (DCA).</p> Results <p>A total of 456 patients were included, with 364 in the training cohort (52 in AVF and 312 in NRF) and 92 in the validation cohort (11 AVF, 81 NRF). Multivariate logistic regression and LASSO regression identified age, T-score, paraspinal fat infiltration (FI), intervertebral disc degeneration (IDD), upper endplate fracture, and bone cement leakage as key predictors for AVF. A risk score system (0–12 points) was developed based on these factors. ROC analysis and calibration curve indicated this scoring system has good discriminatory ability (AUC: 0.874) and goodness-of-fit in the training cohort. At a threshold of ≥ 7 points, the system achieved 88.5% sensitivity and 73.4% specificity. In the validation cohort, this scoring system retained strong discrimination (AUC: 0.826) and good calibration. DCA demonstrated a good clinical net benefit within the probability threshold range of 0.01 to 0.48 in both cohorts.</p> Conclusion <p>This scoring system shows good clinical performance. An intraoperative score ≥ 7 indicates high AVF risk, while preoperative scores ≥ 6 warrant careful bone cement injection to minimize cement leakage.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of risk scoring system for adjacent vertebral fractures after percutaneous vertebroplasty

  • 期刊论文

  • 发表日期：

    2025年11月28日

查看全文
  • Xiao-Yong Ge,
  • Ying-Jie Wang,
  • Shu-Bao Zhang,
  • Chang-Xu Ren,
  • Jin Yang,
  • Ke Xu,
  • Shan-Jin Wang

摘要

Purpose

Patients undergoing percutaneous vertebroplasty (PVP) are at a significant risk of developing adjacent vertebral fractures (AVF) following osteoporotic vertebral fractures (OVFs). Therefore, establishing a risk scoring system for AVF is crucial for timely intervention.

Materials and methods

Patients who underwent PVP for OVFs at our hospital from June 2016 to August 2022 were retrospectively included, with cases from June 2016 to May 2021 used for model training and June 2021 to August 2022 for validation. Patients were divided into AVF and non-refracture (NRF) groups according to whether AVF occurred after treated within 24 months in both cohorts. Potential predictors were compared, and independent risk factors were identified using multivariate logistic and LASSO regression. A risk scoring system was developed and evaluated for discrimination (ROC curve), calibration, and clinical utility (DCA).

Results

A total of 456 patients were included, with 364 in the training cohort (52 in AVF and 312 in NRF) and 92 in the validation cohort (11 AVF, 81 NRF). Multivariate logistic regression and LASSO regression identified age, T-score, paraspinal fat infiltration (FI), intervertebral disc degeneration (IDD), upper endplate fracture, and bone cement leakage as key predictors for AVF. A risk score system (0–12 points) was developed based on these factors. ROC analysis and calibration curve indicated this scoring system has good discriminatory ability (AUC: 0.874) and goodness-of-fit in the training cohort. At a threshold of ≥ 7 points, the system achieved 88.5% sensitivity and 73.4% specificity. In the validation cohort, this scoring system retained strong discrimination (AUC: 0.826) and good calibration. DCA demonstrated a good clinical net benefit within the probability threshold range of 0.01 to 0.48 in both cohorts.

Conclusion

This scoring system shows good clinical performance. An intraoperative score ≥ 7 indicates high AVF risk, while preoperative scores ≥ 6 warrant careful bone cement injection to minimize cement leakage.