Date on Master's Thesis/Doctoral Dissertation


Document Type

Master's Thesis

Degree Name

M. Eng.



Committee Chair

Voor, Michael

Committee Co-Chair (if applicable)

Depuy, Gail

Committee Member

Depuy, Gail

Committee Member

Roussel, Thomas


It is often difficult to achieve adequate screw fixation for plate constructs in fractures with poor quality bone or in metaphyseal bone with extensive bone loss or comminution[7][17]. Furthermore, rescuing or augmenting failed screw fixation in inadequate bone with various cement products has yielded mixed results when tested with pure axial pullout loading[20][21]. In most cases, plate/screw constructs experience both vertical (translational) and horizontal (pullout) forces during physiologic loading[17][20][21]. The increased use of locked screws in plate constructs has also changed the loading patterns of bone screws. For this study, a novel “toggle-out” testing method was developed to more realistically simulate in-vivo loading of screw-plate constructs. Our objective was to compare the fixation of locked and non-locked screws in simulated cancellous bone of three different densities and to determine the effectiveness of augmentation of the screw fixation using either PMMA or a resorbable Calcium Phosphate cement in both stripped and oversized screw holes. Polyurethane foam blocks of 12.5, 20, and 30 lb/ft3 densities representing ostoporotic, normal, and high density cancellous bone respectively (Sawbones, Pacific Research Laboratories, Vashon Island, WA) were used as the bone surrogate for this study. Holes were drilled into the blocks perpendicular to a single face for placement of screws. All screws tested were 4 mm diameter, 32, 34, or 36 mm length stainless steel cancellous bone screws (Stryker, Mahwah, NJ). The holes were either 2.5 mm diameter pilot holes, 4.0 mm diameter (to simulate a stripped screw hole), or 12 mm diameter (bone loss / void). In the 4 mm stripped holes and the12 mm holes, various cements were used to augment the screw fixation. The cements used were PMMA (Simplex, Stryker) and Calcium Phosphate Cement (Trabexus, Vivorté). After placement of the screws or after the cement had set for 24 hours, the blocks were mounted on a load frame (MTS Corp) for cyclic testing. The load fixture allowed screws to be configured either as locked screws or non-locked screws with respect to the plate. Along with cyclic transverse loading, a constant axial pullout force of 20 N was applied to each screw during testing. Cyclic “toggle” loading was applied for 1000 cycles at each of ±25, ±50, ±100, and ±200 N, or until failure by pullout or screw breakage. The average total displacement (positive and negative combined) value for each test was recorded over the last ten load cycles. Under all conditions, the locked screws exhibited significantly less displacement than the non-locked screws (p screws to be set up as locked screws for half of the test conditions. Locking screws exhibited less displacement than non-locking screws across all test samples (P=0.00). This study therefore supports the use of a locking-style cancellous screw in poor quality bone or when cement augmentation of large holes is warranted. The behavior in stripped holes was quite interesting and erratic. The non-locking screws easily pulled out in most cases, but the locking screws were able to survive more cycles. Also, the higher density bone made cement augmentation to prevent pullout more difficult because the cement (especially CaP) does not interdigitate with the high density material. Cement augmentation of large defects and stripped holes in poor quality bone has the potential to be successful regardless of the type of cement used because locking screws were significantly more stable than non-locking screws.