Risk factors and score for recollapse of the augmented vertebrae after percutaneous vertebroplasty in osteoporotic vertebral compression fractures

Abstract

Summary: Our study demonstrated a high incidence of recollapse of the augmented vertebrae after PVP treatment for OVCFs. A risk score based on all significant factors can predict the rate of recollapse and gain clinical benefits to prevent recollapse in patients at high risk. Background: Recollapse of the augmented vertebrae after percutaneous vertebroplasty (PVP) treatment for osteoporotic vertebral compression fractures (OVCFs) has obtained much attention. However, little is known about risk factors and score for recollapse of the augmented vertebrae. Objective: To determine risk factors and furthermore develop a risk score related to recollapse of the augmented vertebrae after PVP treatment for OVCFs. Methods: Patients who were treated with PVP for single OVCFs and met this study’s inclusion criteria were retrospectively reviewed. The follow-up period was at least 2 years. Associations of recollapse with co-variates (age, gender, bone mass density [BMD] with a T-score, fracture level, intravertebral cleft [IVC], fracture type, cement volume, cement leakage, leakage into a disc, cement distribution pattern, Non-PMMA-endplate-contact [NPEC], preoperative fracture severity, reduction rate [RR], reduction angle [RA]) were analyzed and a risk score for recollapse was further developed to predict recollapse. Results: A total of 152 patients were included. Recollapse group was found in 42 (27.6%) patients. Preoperative IVC, solid lump cement distribution pattern, more RR (a cutoff value of 7%) and larger RA (a cutoff value of 3°) was significantly associated with increased risk for recollapse of the augmented vertebrae. A risk score was developed based on the number of risk factors present in each patient. Patients with a score of 4 had an approximately ninefold increased risk of developing recollapse over patients with a score of 0. The receiver operating characteristic curve of the risk score generated an area under the curve of 0.899 (95% CI 0.642–0.836, P = 0.000). Conclusion: A risk score based on preoperative IVC, cement distribution pattern, reduction rate, and reduction angle predicts the rate of recollapse. Additional studies should aim to validate this score and inspect clinical benefits of recollapse prophylaxis in patients at high risk.