Biomechanical comparison of two new atlantoaxial fixation techniques with C1-2 transarticular screw-graft fixation

Two new techniques for atlantoaxial fixation have been recently described. In one technique, C-2 intralaminar screws are connected with C-1 lateral mass screws; in the second, C-1 and C-3 lateral mass screws are interconnected and C-2 is wired sublaminarly. Both techniques include a C1-2 interspinous graft. The authors compared these techniques with the gold-standard, interspinous graft-augmented C1-2 transarticular screw fixation and with a control C1-2 interspinous graft fixation procedure alone.

In six human cadaveric occiput-C4 specimens, nonconstraining 1.5-Nm pure moments were applied to induce flexion, extension, lateral bending, and axial rotation during which three-dimensional angular motion was measured optoelectronically. Each specimen was tested in the normal state, with graft alone (after odontoidectomy), and then in varying order after applying each construct with a rewired graft. All three constructs allowed significantly less angular motion at the C1-2 junction than the wired interspinous graft alone during lateral bending and axial rotation (p < 0.01, paired Student t-test) but not during flexion or extension. Transarticular screw fixation with an interspinous graft allowed less motion at the atlantoaxial junction than the two new constructs in several conditions. Differences were greater between the transarticular screw construct and the intralaminar screw construct than between the transarticular screw construct and the C1-3 lateral mass screw construct. During lateral bending and axial rotation, the C1-3 construct allowed less motion at the atlantoaxial junction than the intralaminar screw construct.

Biomechanically, the gold-standard C1-2 transarticular screw fixation outperformed the two new techniques during lateral bending and axial rotation. Wiring C-2 to C1-3 rods provided greater stability than C1-2 laminar screws, but it sacrificed C2-3 mobility. It is unknown whether the small differences observed biomechanically would lead to clinically relevant differences in fusion rates.