[Biomechanical stability of unilateral pedicle screw fixation on cadaveric model simulated two-level posterior lumbar interbody fusion].

OBJECTIVE: To analyze the biomechanical efficacy of unilateral pedicle screw fixation on human cadaveric lumbar spine model simulated by two-level posterior lumbar interbody fusion (PLIF).

METHODS: Six fresh-frozen adult human cadaveric lumbar spine motion segments (L(2)-S(2)) were simulated to unilateral/bilateral L(4)-S(1) PLIF constructs augmented by unilateral/bilateral pedicle screw fixation sequentially and respectively. All configurations were tested by MTS 858 in the following sequential construct order: the intact, UI (unilateral instability), UIUF1C (unilateral instability via unilateral pedicle screw fixation plus one cage), BIUF1C (bilateral instability via unilateral pedicle screw fixation plus one cage), BIBF1C (bilateral instability via bilateral pedicle screw fixation plus one cage) and BI (bilateral instability without pedicle screw and cage). Each specimen was nondestructively tested in flexion/extension, lateral bending, and axial rotation. An axial compressive load ranged from 40 N to 360 N and the maximum peak moment of 8 N·m was applied during testing. The range of motion (ROM) and neutral zone (NZ) of fusion segment were recorded by a 6-Eagle Motion Analysis F40 system, and then statistic comparison were performed between different simulated constructs with One Way of ANOVA and Post hoc LSD tests.

RESULTS: BIBF1C had the lowest ROM and NZ of L(4)-S(1) fusion segments in all loading models, which were significantly lower than those of any uninstrumented construct (the intact, UI and BI) (P < 0.05). In flexion/extension, lateral bending, and axial rotation, the ROM of UIUF1C was respectively 2.53 ± 1.12, 4.03 ± 2.19, 2.78 ± 1.00 and the NZ of UIUF1C was respectively 1.14 ± 0.70, 1.96 ± 1.13, 1.28 ± 0.71, which were significantly lower than those of the intact (P < 0.05). Compared to BIBF1C, the ROM and NZ were respectively increased 60.13% and 17.52% in flexion/extension, 315.46% and 243.86% in lateral bending, 8.17% and 6.20% in axial rotation, however, there were no significant differences between these two constructs (P > 0.05). In lateral-bending and axial rotation, the ROM and NZ of BIUF1C were significantly higher than those of BIBF1C (P < 0.05). In flexion/extension, the ROM and NZ of BIUF1C were higher than those of BIBF1C but there were no significant differences (P > 0.05). Compared to the intact, BIUF1C had lower ROM and NZ except for higher NZ in axial rotation, and there were significant differences only in flexion/extension (P < 0.05).

CONCLUSIONS: All tested two-level unilateral fixation on simulated human cadaveric model with unilateral PLIF can achieve similar initial biomechanical stability in comparison with two-level bilateral pedicle screw fixation. However in most test modes, two-level unilateral pedicle screw fixation on simulated human cadaveric model with bilateral PLIF can not achieve enough biomechanical efficacy in comparison with two-level bilateral pedicle screw fixation.