Predicting femoral neck screw pullout strength using DXA and novel assessments of bone microarchitecture

Abstract

Abstract Summary This study determined how well DXA and other novel assessments of bone microarchitecture predicted screw pullout strength in the femoral neck. Hip-specific DXA performed as well as the novel approaches at predicting screw pullout. These findings support the use of DXA to evaluate bone quality and aid in surgical decision-making. Purpose The primary purpose of our study was to determine how well dual-energy x-ray absorptiometry (DXA), trabecular bone score (TBS Ortho), 3D-Shaper, and quantitative computed tomography (qCT) predicted screw pullout of cannulated screws placed in the femoral neck of cadaveric specimens. Our secondary objective was to determine how well insertional torque predicted screw pullout. Methods We conducted a biomechanical study using twenty proximal cadaveric femurs. Each femur underwent imaging with each respective modality to assess bone quality. Screws were placed in a standardized configuration, and insertional torque was measured using a digital torque screwdriver. Screws were then pulled out using a mechanical testing machine to determine peak pullout strength. The relationship between screw pullout force and each bone metric was assessed using Spearman correlation coefficients and univariate linear mixed-effects models. Results DXA bone mineral density (BMD) and 3D-Shaper volumetric BMD (vBMD) at the femoral neck had the highest correlation coefficients with screw pullout strength ( r = 0.95, p < 0.001). However, all of the imaging modalities were found to be strong predictors of screw pullout. Insertional torque was also a strong predictor of screw pullout strength ( r = .86, p < 0.001). Conclusion Our findings highlight the clinical utility of hip-specific DXA as a strong and accessible predictor of screw pullout strength, performing as well as newer imaging modalities such as TBS Ortho and 3D-Shaper. Insertional torque also demonstrated strong predictive value and may serve as a useful intraoperative tool when preoperative bone density data are not available.