Residual motion on flexion-extension radiographs after simulated lumbar arthrodesis in human cadavers

Flexion-extension radiographs are commonly used to assess lumbar fusion. Recommended criteria for solid fusion have varied from 1 to 5 degrees of angular motion between vertebrae. Notwithstanding this wide variation, the validity of these criteria have never been biomechanically tested. As a preliminary and initial step, it was the authors' purpose to quantify measurable angular motion after simulating solid lumbar fusion in human cadaver spines. Seven cadaveric spines (L1 to L4) were tested in a radiolucent jig fixed to a servohydraulic testing apparatus. Flexion and extension moments of 10 Nm were applied. Fusion was simulated using metallic implants spanning the L2-L3 motion segment. These included transverse process plates, a spinous process plate, pedicle screw construct, or an anterior vertebral body plate to simulate an intertransverse, interspinous process, facet, and interbody fusions, respectively. Angular movements were measured on lateral radiographs and statistically compared using a repeated measures analysis of variance. Simulated intertransverse fusion resulted in 13+/-4 degrees of motion; interspinous fusion, 9+/-4 degrees; posterior facet fusion, 5+/-3 degrees; and interbody fusion with plate, 3+/-2 degrees. Compared with the intact, only posterior facet fusion and interbody fusion with plate had statistically significantly less motion (P=0.006 and 0.0001, respectively). The amount of radiographically detectable flexion-extension motion with simulated fusions varies widely and seems to be influenced by fusion type. This study documents a range of measurable motion on flexion-extension radiographs after several types of simulated lumbar fusion. However, as the degrees of motion seemed to be high, future studies should use a fusion simulation other than metallic implants that more closely resembles bony arthrodesis.