Risk factors for local bone destruction progression in palliative percutaneous vertebroplasty for vertebral metastases and the significance of bone cement filling rates

Abstract

Background: For palliative percutaneous vertebroplasty (PVP) for vertebral metastases, local bone destruction progression (LBDP) commonly occurs in the previously treated vertebrae. There were no studies regarding LBDP and its risk factors in previous reports, and there was no uniform evaluation method for the distribution of bone cement in the vertebrae. Objectives: We aimed to investigate the risk factors for LBDP after PVP for palliative treatments in patients with vertebral metastases. We also proposed that filling rates could be used as a simple evaluation method to detect vertebral metastases and explored its clinical significance. Study Design: This was a retrospective study. Setting: A university hospital. Methods: A total of 48 patients and 54 vertebrae that had received PVP as a palliative treatment for vertebral metastases were recruited between October 2012 to October 2019 from the Shengjing Hospital of the China Medical University. We collected and evaluated the data including age, gender, cement filled completely or not, cement dose used, the cement distribution score, time of LBDP, and so on, and the filing rate we proposed was also included. Results: This retrospective study divided 48 patients and 54 vertebrae into group A for those with an LBDP of less than 6 months (n = 41), and group B for those with an LBDP of 6 or more months (n = 13). The complete filling of bone cement and bone cement dose in group B was much higher than that in group A (2.85 ± 0.97 vs. 4.12 ± 1.77; P = 0.027), and the time of recurrent pain in group B was significantly higher compared with that in group A (8.46 ± 2.73 vs. 3.39 ± 1.63; P < 0.0001). There was a statistical difference in the Saliou score and filling rate between the 2 groups (11.77 ± 3.17 vs. 9.34 ± 3.28, P = 0.023; 0.752 ± 0.227 vs. 0.489 ± 0.161, P < 0.0001). Univariate logistic analysis showed that complete filling of cement, the cement dose, Saliou score, and filling rate were statistically significant predictors of LBDP occurring in less than 6 months. Multivariate logistic analysis showed that the filling rate was an independent predictor of patients with vertebral metastases developing LBDP in less than 6 months (odds ratio, < 0.001; 95% confidence interval, < 0.001-0.006; P = 0.0007). The cutoff value for the filling rate calculated from the receiver operating characteristic (ROC) curve analysis was 0.646, which could identify patients who had LBDP in less than 6 months of PVP with a sensitivity of 85.4% and specificity of 84.6%. The 6-month LBDP in the 0.646 or less ROC curve group was higher than that in the greater than 0.646 ROC curve group (97.22% vs. 55.56%, P < 0.0001). Limitations: The retrospective nature and small sample size were significant. Variation in the time and state of bone cement injected during all PVP procedures was a bias. There was no pathological diagnosis of all vertebral metastases. Conclusions: The cement dose, complete filling of cement, Saliou score, and filling rate were factors negatively related to LBDP occurring in less than 6 months. Patients with lower filling rates are maybe more likely to have early LBDP compared with those with higher filling rates.