Evaluating Atlantoaxial Dislocation Based on Cartesian Coordinates: Proposing a New Definition and Its Impact on Assessment of Congenital Torticollis.

Background: Conventional 2-dimensional (2-D) definition of atlantoaxial dislocation (AAD) is inadequate for coexisting 3-D displacements.

Objective: To prospectively classify AAD and its related abnormalities along 3 Cartesian coordinates and assess their association with torticollis.

Methods: One hundred and fifty-four patients with congenital AAD were prospectively classified according to their C1-2 displacement along 3 Cartesian coordinates utilizing 3-D multiplanar CT. The impact of this 3-D dislocation on occurrence of clinically manifest torticollis was also evaluated and surgical treatment was planned.

Results: Three dimensional CT assessment detected the following types of C1-2 dislocations: I:translational dislocation (along Z coordinate, n = 37 [24%]); II: central dislocation (along Y coordinate, n = 10 [6.5%]); III: translational+central dislocation (along Z+Y coordinates, n = 42 [27.3%]); IV: translational dislocation+ rotational dislocation+coronal tilt (along Z+X coordinates, (n = 6 [3.9%]); V: central dislocation (basilar invagination)+rotational dislocation+coronal tilt (along Y+X coordinates, n = 11 [7.1%]); VI: translational dislocation+ central dislocation+ rotational dislocation+ coronal tilt (along all 3 axes, n = 48 [31%]). Assessing degree of relative C1-2 rotation revealed that 27 (37%) of 85 patients with <50 rotation and 54 (78%) of 69 patients with >5° rotation had associated torticollis. Translational dislocation had negative association (odds ratio [OR] 0.1, 95% confidence interval [CI; 0.47-0.32], P = .00), while type VI (OR 5.0, 95% CI [2.2-11.19], P = .00), type V (OR 4.44, 95% CI [0.93-21.26], P = .04), and type IV (OR 1.84, 95% CI [0.32-10.38], P = .48) dislocations had strong positive association with torticollis. Sixty-two (40%) patients improved, 68 (44%) remained unchanged, and 24 (16%) patients worsened postoperatively. Twenty-eight patients required second-stage transoral decompression following posterior distraction-fusion due to neurological nonimprovement.

Conclusion: Three-dimensional assessment of AAD including evaluation of culpable C1-2 facet joints addresses anomalous displacements in 3 Cartesian planes. This provides targets for adequate cervicomedullary decompression-stabilization, and helps in the management of accompanying torticollis.