Biomechanical characteristics of pedicle screws in osteoporotic vertebrae-comparing a new cadaver corpectomy model and pure pull-out testing.

Currently, evaluation of the stability of spinal instrumentations often focuses on simple pull-out or cyclic loading. However, the loading characteristics and the specimen alignment rarely simulate physiological loading conditions, or the clinical situation itself. The purpose of this study was to develop an alternative setup and parameters to compare static and dynamic characteristics of pedicle screws at the bone-implant interface in lumbar osteoporotic cadavers. A corpectomy model development was based on ASTM-1717 standard, allowing a deflection of the cranial and caudal element under loading. Twelve human osteoporotic vertebrae (L1-L4) were analyzed for morphological CT-data and T-Score. For group A (n = 6) loads were simulated as in vivo measurements during walking, representing 2 months postoperatively. A subsequent pull-out was performed. Group B (n = 6) was tested with pure pull-out. Screw loosening at the tip/head was optically measured and analyzed with respect to clinical patterns. Correlations between CT-data, T-Score, and in vitro parameters were determined. For group A, the subsidence for the head/tip was measured towards the upper/lower endplate, resulting in visible deflections. The progress of the subsidence was greatest within the first and last cycles until failure. The predominant patterns were pure rotation and toggling. However, the pull-out between groups was not significantly different. Pedicle-angle and cyclic-subsidence correlated with R = 0.806/0.794. T-Score and pull-out correlated only in group A. With the corpectomy setup, clinically observed wipe effects and a loss of correction could be simulated. The presented parameters facilitate analysis of the complex changing load distributions and interactions between the left and right bone-implant interface.