Purpose Pedicle screw pull-out remains a challenge in thoracolumbar spine fixation, contributing to fixation failure. Computational simulations offer a pathway to optimize screw designs across diverse materials and bone qualities towards reducing failure rates. Non-linear explicit micro-finite element (μFE) simulation was hypothesized to accurately predict pullout forces of CF/PEEK screws, with bone volume fraction (BV/TV) influencing performance. This study aimed to validate the μFE model and assess BV/TV effects on the basis of experimental testing.Methods Thirteen cadaveric vertebrae were instrumented with 4.5 mm (n = 13) and 5.5 mm (n = 9) CF/PEEK screws using sample-specific 3D-printed guides. Quasi-static pullout tests were conducted and simulated using non-linear explicit μFE models based on high-resolution peripheral quantitative computed tomography (HR-pQCT). Regression analysis evaluated the relationship between BV/TV and pullout force. Correlation coefficient (R2), concordance correlation coefficient (CCC), standard error of estimate (SEE), and relative standard error (RSE) were used to assess the agreement between experimental and μFE-predicted pullout forces.Results The linear regression relationship between BV/TV and pullout force was significantly different for the 4.5 mm and 5.5 mm screws (p = 0.001). Ordinary least squares (OLS) regression showed significant BV/TV influence on pullout strength (p = 0.001 experimental, p < 0.001 μFE). Linear regression in log10-log10 space for experimental versus μFE forces showed strong correlation (R2 = 0.931, p < 0.001), despite μFE overprediction (slope = 0.752 vs. 1:1, p < 0.001; CCC = 0.745, SEE = 428.7 N, RSE = 85.2 %).Conclusions The validated μFE model accurately predicted pullout strength of pedicle screw of different designs, supported by strong correlation with experimental data, despite light overprediction, potentially due to the lack of insertion damage modeling. The robust in silico framework can be used to enhance orthopedic screw designs, supporting improved fixation stability across diverse bone qualities.
