Which postures are most suitable in assessing spinal fusion using radiostereometric analysis?

BACKGROUND Up to now, plain radiographs are not well suited to assess spinal fusion. Radiostereometric analysis performed for two postures may deliver more reliable results. However, it is unknown, which postures are most suitable for this procedure. METHODS In a finite element study, spinal fusion at the level L4-5 was simulated assuming a posterior approach and the implantation of two cages and a spinal fixation device. The change of the distance between markers in vertebrae adjacent to the cages was calculated for moving from one of the following postures standing, flexion, extension, axial rotation, lying, and extension in a lying position to another. The changes of marker distances were calculated for the intact model, as well as for the situations: directly after surgery before fusion started, in the early-fusion-phase and in the late-fusion-phase. Differences in the marker motion between two postoperative situations were also calculated. FINDINGS The most anteriorly placed markers showed the greatest motion between two postures. The greatest differences in marker motions between the two situations before-fusion and early-fusion-phase (0.54 mm) as well as between early-fusion-phase and late-fusion-phase (0.34 mm) were found for the two postures flexion while standing and extension in a lying position. INTERPRETATION Pairs of X-rays taken while standing with maximum flexed upper body and while lying with maximum extended trunk are most suited for the assessment of spinal fusion when using radiostereometric analysis.