Abstract
Low bone mineral density (BMD) impairs the stability of the bone-screw interface, which leads to screw loosening after spinal instrumentation. Quantitative computed tomography (QCT) and dual-energy X-ray absorptiometry (DXA) were applied to measure BMD of screw trajectories and other regions of the vertebrae in this paper, and the aim was to analyze the best effective tool and BMD of the best appropriate vertebral site to predict pedicle screw loosening after lumbar fusion surgery. 186 patients who underwent lumbar interbody fusion and pedicle screws placement were analyzed retrospectively. Patients were divided into 2 groups according to whether there was screw loosening: fissure is greater than or equal to 1 mm around the screws at follow-up CT scans. The volumetric BMD (vBMD) was measured by QCT in the central vertebral body, pedicle, and screw trajectory region, and DXA was applied for the lumbar spine and hip area BMD (aBMD). The overall pedicle screw loosening rate was 33.9% (63/186). Demographic data, health history, and the lumbar aBMD were not significantly different between the two groups. Multivariate analysis revealed showed that the hip aBMD, vBMD in the central vertebral body, pedicle, and screw trajectory regions were independent risk factors for screw loosening. Additionally, Receiver operating characteristic curve revealed the screw trajectory vBMD had the greatest area under the curve for predicting screw loosening. The screw trajectory vBMD using QCT had a stronger predictive value than the vBMD in other regions of the vertebrae and the hip aBMD, and had a more representative bone quality measurement in the bone-screw binding region.