Effects of multilevel posterior ligament dissection after spinal instrumentation on adjacent segment biomechanics as a potential risk factor for proximal junctional kyphosis: a biomechanical study

BACKGROUND Spinous processes and posterior ligaments, such as inter- and supraspinous ligaments are often sacrificed either deliberately to harvest osseous material for final spondylodesis e.g. in deformity corrective surgery or accidentally after posterior spinal instrumentation. This biomechanical study evaluates the potential destabilizing effect of a progressive dissection of the posterior ligaments (PL) after instrumented spinal fusion as a potential risk factor for proximal junctional kyphosis (PJK). METHODS Twelve calf lumbar spines were instrumented from L3 to L6 (L3 = upper instrumented vertebra, UIV) and randomly assigned to one of the two study groups (dissection vs. control group). The specimens in the dissection group underwent progressive PL dissection, followed by cyclic flexion motion (250 cycles, moment: + 2.5 to + 20.0 Nm) to simulate physical activity and range of motion (ROM) testing of each segment with pure moments of ±15.0 Nm after each dissection step. The segmental ROM in flexion and extension was measured. The control group underwent the same loading and ROM testing protocol, but without PL dissection. RESULTS In the treatment group, the normalized mean ROM at L2-L3 (direct adjacent segment of interest, UIV/UIV + 1, PJK-level) increased to 104.7%, 107.3%, and 119.4% after dissection of the PL L4-L6, L3-L6, and L2-L6, respectively. In the control group the mean ROM increased only to 103.2%, 106.7%, and 108.7%. The ROM difference at L2-L3 with regard to the last dissection of the PL was statistically significant (P = 0.017) and a PL dissection in the instrumented segments showed a positive trend towards an increased ROM at UIV/UIV + 1. CONCLUSIONS A dissection of the PL at UIV/UIV + 1 leads to a significant increase in ROM at this level which can be considered to be a risk factor for PJK and should be definitely avoided during surgery. However, a dissection of the posterior ligaments within the instrumented segments while preserving the ligaments at UIV/UIV + 1 leads to a slight but not significant increase in ROM in the adjacent cranial segment UIV/UIV + 1 in the used experimental setup. Using this experimental setup we could not confirm our initial hypothesis that the posterior ligaments within a long posterior instrumentation should be preserved.