Biomechanical Investigation of Cortical Screws in the Proximal Thoracic Spine: The Effects of Bone Density, Screw Insertion Torque, Screw Trajectory, and Tapping in Resisting Catastrophic Hardware Failure

Presented at SMISS Annual Forum 2016
By Jonathan Harris MS
With Sun-Ren Sheng MD, PhD, Mir Hussain MS, Hua-Zi Xu MD, Mariano Titanti BS, Xin Fu MD, Brandon Bucklen PhD, Bryan Cunningham PhD

Disclosures: Jonathan Harris MS E; Globus Medical, Inc., Sun-Ren Sheng MD, PhD None, Mir Hussain MS E; Globus Medical., Hua-Zi Xu MD None, Mariano Titanti BS None, Xin Fu MD None, Brandon Bucklen PhD E; Globus Medical Inc., Bryan Cunningham PhD E; Globus Medical, Inc.

Introduction

Posterior rods and pedicle screws (PS) remain the gold standard in restoring spinal alignment for thoracospinal deformities. Cortical screws (CS) are a recent development in minimally invasive surgery (MIS) technology, and considered muscle-sparing compared to PS. While resistance of catastrophic failure/pullout has been quantified in the lumbar spine, performance of CS in the upper thoracic spine and effect of pedicle tapping remains unknown. 

Aims/Objectives

Determine whether PS and CS have equivalent pullout strength following simulated in vitro fatiguing within a human thoracic cadaveric model. 

Methods

Twelve specimens (T2-T9) were divided in four groups (PS and CS, with and without tapping) such that bone mineral density was equivalent between groups. Specimens were potted at T2 and T9 (Bondo, Bondo Corp, Atlanta, GA). Operative segments included T3-T4, T5-T6. Posterior rods and screws consisted of titanium rods and either 4.0/5.0mm CS (25-30mm length) or 4.5mm (30-40mm length) PS (Globus Medical, Inc., Audubon, PA). Specimens experienced cyclic loading of 7.5Nm for 3000 cycles in flexion-extension (FE), lateral bending (LB), and axial rotation (AR) to simulate long term fixation. Screw insertion torque and ultimate load-to-failure of the bone-screw interface was recorded. Multiple regression analysis, independent and paired t-tests were used to discern influence of BMD, use of tapping, screw trajectory, and screw insertion torque on pullout strength (significance: p<0.05). 

Results

Screw insertion torques for CStap, CSno-tap, PStap, and PSno-tap were 2.0±0.7Nm, 2.5±1.4Nm, 1.6±0.5Nm, and 1.3±0.5Nm respectively. Pullout forces for CStap, CSno-tap, PStap, and PSno-tap were 590±300Nm , 830±430Nm, 854±380N, and 500±260N respectively. BMD, screw trajectory, and insertion torque significantly predicted screw pullout strength, F(4, 138) = 47.17, p < 0.01, R2 = 0.565 (p<0.05). CS insertion torque was higher than PS, regardless of tapping (p<0.05). No differences were found between PStap and CSno-tap (p>0.05).

Conclusions

Pedicle fixation techniques in the thoracic spine may be nuanced. Tapping of the pedicle reduced pullout strength of cortical screws, while the opposite was observed for the pedicle screw. Results suggest tapping cortical bone compromises the cortical shell while tapping along the midline of the pedicle compacts adjacent cancellous bone, thus increasing pullout resistance of pedicle screws. Cortical screws, without tapping, provides equivalent resistance to failure as pedicle screws with tapping. Clinical investigation is required to understand impact of the results.