Predictive Factors for Bone Cement Displacement following Percutaneous Vertebral Augmentation in Kümmell's Disease

Objective: To investigate the independent influencing factors of bone cement displacement following percutaneous vertebral augmentation (PVA) in patients with stage I and stage II Kümmell’s disease. Methods: We retrospectively reviewed the records of 824 patients with stage Ⅰ and stage Ⅱ Kümmell’s disease treated with percutaneous vertebroplasty (PVP) or percutaneous vertebroplasty (PKP) from January 2016 to June 2022. Patients were divided into the postoperative bone cement displacement group (n = 150) and the bone cement non-displacement group (n = 674) according to the radiographic inspection results. The following data were collected: age, gender, body mass index (BMI), underlying disease, bone mineral density (BMD), involved vertebral segment, Kümmell’s disease staging, anterior height, local Cobb angle, the integrity of anterior vertebral cortex, the integrity of endplate in surgical vertebrae, surgical method, surgical approach, the volume of cement, distribution of cement, the viscosity of cement, cement leakage, and postoperative anti-osteoporosis treatment. Binary logistic regression analysis was performed to determine the independent influencing factors of bone cement displacement. The discrimination ability was evaluated using the area under the curve (AUC) of the receiver operating characteristic (ROC). Results: The results of logistic regression analysis revealed that thoracolumbar junction (odds ratio (OR) = 3.23, 95% confidence interval (CI) 2.12−4.50, p = 0.011), Kümmell’s disease staging (OR = 2.23, 95% CI 1.81−3.41, p < 0.001), anterior cortex defect (OR = 5.34, 95% CI 3.53−7.21, p < 0.001), vertebral endplates defect (OR = 0.54, 95% CI 0.35−0.71, p < 0.001), cement distribution (OR = 2.86, 95% CI 2.03−3.52, p = 0.002), cement leakage (OR = 4.59, 95% CI 3.85−5.72, p < 0.001), restoration of local Cobb angle (OR = 3.17, 95% CI 2.40−5.73, p = 0.024), and postoperative anti-osteoporosis treatment (OR = 0.48, 95% CI 0.18−0.72, p = 0.025) were independently associated with the bone cement displacement. The results of the ROC curve analysis showed that the AUC was 0.816 (95% CI 0.747−0.885), the sensitivity was 0.717, and the specificity was 0.793. Conclusion: Thoracolumbar fracture, stage Ⅱ Kümmell’s disease, anterior cortex defect, uneven cement distribution, cement leakage, and high restoration of the local Cobb angle were risk factors for cement displacement after PVA in Kümmell’s disease, while vertebral endplates defect and postoperative anti-osteoporosis treatment are protective factors.