PURPOSE: To determine the risk factors for new vertebral compression fractures (VCFs) following percutaneous vertebroplasty (PV) in patients with osteoporosis. MATERIALS AND METHODS: This prospective study included 132 consecutive patients with osteoporosis treated with PV in a single institution over 46 months from March 2005 to December 2008. Multivariable logistic regression and univariate analysis were employed to identify risk factors for new VCFs after PV, including patient demographic data, parameters of the initial and new fractured vertebrae, procedure-related information, and follow-up data. RESULTS: During the follow-up period (22.4 months ± 12.1), 80 new vertebral fractures occurred in 36 (27.3%) patients. Multivariate analysis showed that number of VCFs per time frame, computed tomography (CT) value of nonfractured vertebrae (T11-L2), activity level after discharge, duration of follow-up, and cement distribution in the inferior part of the vertebral body or close to the endplate were statistically correlated with new fractures (odds ratios, 2.63, 0.96, 3.59, 1.00, 0.30, and 0.05; P = .006, P = .001, P = .007, P = .004, P = .021 and P = .029). Univariate analysis showed preexisting old VCFs were correlated with new VCFs (P = .045). Subsequent compression fractures in adjacent vertebrae (45 of 80) occurred more frequently and sooner than nonadjacent vertebral fractures (both P < .05). CONCLUSIONS: The incidence of new VCFs after PV is relatively high and affected by several risk factors that are related to both the PV procedure and the natural course of osteoporosis.
PURPOSE: To determine the risk factors for new vertebral compression fractures (VCFs) following percutaneous vertebroplasty (PV) in patients with osteoporosis. MATERIALS AND METHODS: This prospective study included 132 consecutive patients with osteoporosis treated with PV in a single institution over 46 months from March 2005 to December 2008. Multivariable logistic regression and univariate analysis were employed to identify risk factors for new VCFs after PV, including patient demographic data, parameters of the initial and new fractured vertebrae, procedure-related information, and follow-up data. RESULTS: During the follow-up period (22.4 months ± 12.1), 80 new vertebral fractures occurred in 36 (27.3%) patients. Multivariate analysis showed that number of VCFs per time frame, computed tomography (CT) value of nonfractured vertebrae (T11-L2), activity level after discharge, duration of follow-up, and cement distribution in the inferior part of the vertebral body or close to the endplate were statistically correlated with new fractures (odds ratios, 2.63, 0.96, 3.59, 1.00, 0.30, and 0.05; P = .006, P = .001, P = .007, P = .004, P = .021 and P = .029). Univariate analysis showed preexisting old VCFs were correlated with new VCFs (P = .045). Subsequent compression fractures in adjacent vertebrae (45 of 80) occurred more frequently and sooner than nonadjacent vertebral fractures (both P < .05). CONCLUSIONS: The incidence of new VCFs after PV is relatively high and affected by several risk factors that are related to both the PV procedure and the natural course of osteoporosis.