Impact of constrained dual-screw anchorage on holding strength and the resistance to cyclic loading in anterior spinal deformity surgery: a comparative biomechanical study.

Study Design: Biomechanical in vitro laboratory study.
Objective: To compare the biomechanical performance of 3 fixation concepts used for anterior instrumented scoliosis correction and fusion (AISF).
Summary of Background Data: AISF is an ideal estimate for selective fusion in adolescent idiopathic scoliosis. Correction is mediated using rods and screws anchored in the vertebral bodies. Application of large correction forces can promote early weakening of the implant-vertebra interfaces, with potential postoperative loss of correction, implant dislodgment, and nonunion. Therefore, improvement of screw-rod anchorage characteristics with AISF is valuable.
Methods: A total of 111 thoracolumbar vertebrae harvested from 7 human spines completed a testing protocol. Age of specimens was 62.9 ± 8.2 years. Vertebrae were potted in polymethylmethacrylate and instrumented using 3 different devices with identical screw length and unicortical fixation: single constrained screw fixation (SC fixation), nonconstrained dual-screw fixation (DNS fixation), and constrained dual-screw fixation (DC fixation) resembling a novel implant type. Mechanical testing of each implant-vertebra unit using cyclic loading and pullout tests were performed after stress tests were applied mimicking surgical maneuvers during AISF. Test order was as follows: (1) preload test 1 simulating screw-rod locking and cantilever forces; (2) preload test 2 simulating compression/distraction maneuver; (3) cyclic loading tests with implant-vertebra unit subjected to stepwise increased cyclic loading (maximum: 200 N) protocol with 1000 cycles at 2 Hz, tests were aborted if displacement greater than 2 mm occurred before reaching 1000 cycles; and (4) coaxial pullout tests at a pullout rate of 5 mm/min. With each test, the mode of failure, that is, shear versus fracture, was noted as well as the ultimate load to failure (N), number of implant-vertebra units surpassing 1000 cycles, and number of cycles and related loads applied.
Results: Thirty-three percent of vertebrae surpassed 1000 cycles, 38% in the SC group, 19% in the DNS group, and 43% in the DC group. The difference between the DC group and the DNS group yielded significance (P = 0.04). For vertebrae not surpassing 1000 cycles, the number of cycles at implant displacement greater than 2 mm in the SC group was 648.7 ± 280.2 cycles, in the DNS group was 478.8 ± 219.0 cycles, and in the DC group was 699.5 ± 150.6 cycles. Differences between the SC group and the DNS group were significant (P = 0.008) as between the DC group and the DNS group (P = 0.0009). Load to failure in the SC group was 444.3 ± 302 N, in the DNS group was 527.7 ± 273 N, and in the DC group was 664.4 ± 371.5 N. The DC group outperformed the other constructs. The difference between the SC group and the DNS group failed significance (P = 0.25), whereas there was a significant difference between the SC group and the DC group (P = 0.003). The DC group showed a strong trend toward increased load to failure compared with the DNS group but without significance (P = 0.067). Surpassing 1000 cycles had a significant impact on the maximum load to failure in the SC group (P = 0.0001) and in the DNS group (P = 0.01) but not in the DC group (P = 0.2), which had the highest number of vertebrae surpassing 1000 cycles.
Conclusion: Constrained dual-screw fixation characteristics in modern AISF implants can improve resistance to cyclic loading and pullout forces. DC constructs bear the potential to reduce the mechanical shortcomings of AISF.