Risk factors for augmented vertebrae recollapse after percutaneous kyphoplasty for osteoporotic vertebral compression fractures

Objectives: The purpose of this study was to analyze the risk factors for recollapse of augmented vertebrae after percutaneous kyphoplasty (PKP). Methods: Patients who had been treated with bilateral PKP and met the inclusion criteria were retrospectively reviewed. We assessed the following potential risk factors for augmented vertebrae recollapse: age, gender, weight, height, body mass index (BMI), preoperative T-score in bone mineral density (BMD), fracture spine level, follow-up duration, reduction rate (RR), reduction angle (RA), cement volume, preoperative intravertebral cleft (IVC), cement contact with endplates, cement distribution, and the cross-sectional area (CSA) and percent fatty infiltration area (pFIA) of the multifidus. Univariate and multivariate regression analyses were progressively performed to identify the risk factors for augmented vertebrae recollapse. Results: A total of 112 patients were enrolled in the study. There were 27 (24.1%) patients in the recollapse group. Multivariate regression analysis identified preoperative IVC, separated cement distribution, higher RR, greater RA, and larger pFIA of the multifidus as the risk factors for the augmented vertebrae recollapse.Larger CSA of the multifidus was a protective factor for vertebral body recompression. Receiver operating characteristic (ROC) curve analysis showed that the areas under the ROC curve of higher RR, greater RA, larger pFIA, and larger CSA were 0.710, 0.649, 0.714,and 0.754, respectively. Conclusions: Preoperative IVC, separated cement distribution, higher RR, greater RA, and larger FIA of the multifidus are independent risk factors for the augmented vertebrae recollapse. Larger CSA of the multifidus is a protective factor for the augmented vertebrae recollapse.