Correlation of pull-out strength of cement-augmented pedicle screws with CT-volumetric measurement of cement

Background: Cement augmentation of pedicle screws increases fixation strength in an osteoporotic spine. This study was designed to determine the cement distribution and the correlation between the pull-out strength of the augmented screw and the cement volume within polyurethane (PU) foam. Methods: Twenty-eight cannulated pedicle screws (6 x 45 mm) (Peter Brehm, Erlangen, Germany) with four holes at the distal end of the screw were augmented with the acrylic Stabilit ER Bone Cement Vertebral Augmentation System (DFine Inc., San Jose, CA, USA) and implanted into open-cell rigid PU foam (Pacific Research Laboratories, Vashon Island, WA, USA) with a density of 0.12 g/cm(3), resembling severe osteoporosis. Volumetric measurement of the cement with consideration of the distribution around the screws was done with multislice computed tomography scan (Somatom Definition, Siemens, Erlangen, Germany). Pull-out strength was tested with a servohydraulic system (MTS System Corporation, Eden Prairie, MN, USA), and nonaugmented screws served as control. Pearson ' s correlation coefficient with significance level a = 0.05 and one-way analysis of variance test were used. Results: We found a high (r = 0.88) and significant (p < 0.01) correlation between the cement volume and the pull-out strength, which increased by more than 5-fold with a volume of 3 ml. The correlation appeared linear at least up to 4 ml cement volume and failure always occurred at the cement-bone interface. The cement distribution was symmetric and circular around the most proximal hole, with a distance of 14 mm from the tip, and nearly 90% of the cement was found 6 mm distal and cranial to it. The 95% confidence interval for the relative amount of cement was 37% -41% within 2 mm of the most proximal hole. Conclusion: Compared with the control, a cement volume between 2.0 and 3.0 ml increased the pull-out strength significantly and is relevant for clinical purposes, whereas a volume of 0.5 ml did not. A cement volume beyond 3.0 ml should further increase the pull-out strength because the correlation was linear at least up to 4.0 ml, but the possibility of in vivo cement leakage with increasing volume has to be considered. Pressure-controlled cement application might be a tool to avoid this complication. The cement almost completely penetrated the most proximal perforation.