A Biomechanical Investigation of Fusionless Anterior Dual Screw and Cord Constructs for Scoliosis Correction

Introduction
Adolescent idiopathic scoliosis (AIS) is traditionally treated by posterior spinal fusion (PSF). PSF is associated with the development of adjacent-level disc degeneration, decreased range of motion (ROM), and decreased functional spinal mobility. Vertebral body tethering (VBT) is an alternative to PSF for curve correction in adolescent scoliosis that uses growth modulation to straighten the spine, with the goal to avoid or postpone fusion. Cord breakage is a potential issue that surgeons can encounter that can mitigated via thicker cords or dual cord constructs.

Aims/Objectives
To evaluate the motion-preserving properties of anterior vertebral body tethering with varying cord/screw constructs and cord thicknesses in cadaveric thoracolumbar spines.

Methods
Six fresh-frozen human cadaveric spine specimens (T1-L5) with full rib cages were selected based on similar age and DEXA scores. Specimens with previous spinal trauma, malignancy, deformity, or fractures were excluded. In vitro flexibility tests were performed on intact specimens and range-of-motion (ROM) was measured. In the thoracic spine (T5-12), two staples and screws were placed in each vertebral body, parallel to endplates, through rib ports on the specimen’s right side. Screws were cemented to mitigate screw loosening during flexibility testing. Staples and cemented screws were then placed in the lumbar spine (T12-L4) on the specimen’s left. Single (4.0mm and 5.0mm diameter) and dual (4.0mm diameter) cords were then sequentially applied to the thoracic spine, tensioned (100N), and tested. This process was also applied to the lumbar spine. All single-cord constructs were fixed to the more posterior vertebral body screws; in the dual cord constructs, one cord was placed in the anterior screw and one in the posterior screw. ROM of intact and treated specimens in flexion-extension (FE), lateral bending (LB), and axial rotation (AR) were measured using a custom-built motion simulator (±7.5Nm), capture system, and analysis software and normalized as a percentage of intact.

Results
In thoracic constructs, single 4.0 and 5.0 cords demonstrated ROM comparable to or greater than intact in FE (Single 4.0: 106.5%, p>0.99; Single 5.0: 105.8%, p>0.99), LB (Single 4.0: 98.5%, p>0.99; Single 5.0: 96.6%, p>0.99), and AR (Single 4.0: 108.2%, p>0.99; Single 5.0: 108.9%, p>0.99), while dual 4.0 cords showed the lowest ROM in FE (92.2%, p>0.99 vs intact), LB (77.8%, p>0.99 vs intact), and AR (101.8%, p>0.99 vs intact). In lumbar constructs, single 4.0 and 5.0 cords demonstrated ROM comparable to intact in FE (Single 4.0: 94.4%, p>0.99; Single 5.0: 97.9%, p>0.99) and less ROM than intact in LB (Single 4.0: 34.6%, p< 0.01; Single 5.0: 41.9%, p=0.01) and AR (Single 4.0: 76.0%, p=0.16; Single 5.0: 77.7%, p=0.21). Dual 4.0 cords showed the lowest ROM in FE (78.9%, p=0.13 vs intact), LB (26.8%, p< 0.01 vs intact), and AR (68.7%, p=0.13 vs intact).

Conclusions
Stiffness was similar between single 4.0 and 5.0 cords and greatest in dual 4.0 cords. Single 4.0 and 5.0 cords had similar stiffness to intact in the thoracic spine in all bending moments, and in the lumbar spine during FE. LB was the most restricted by tethering in thoracic and lumbar constructs.