Atlantoaxial Rotatory Fixation–Subluxation Revisited

Abstract

Study Design. Cross-sectional clinical and radiologic study with a normal control group. Objectives. To compare the range of motion of the atlantoaxial joint in patients with acute torticollis with those of normals as measured from computed tomography scans, to look for the existence of atlantoaxial rotatory fixation in any position (subluxation or normal range of motion) in this group of patients, and to clarify the definition of atlantoaxial rotatory subluxation by measuring the atlantodental interval and analyzing the location of the center of rotation in patients as well as normal controls. Summary of Background Data. Although acute acquired torticollis is usually termed atlantoaxial rotatory subluxation or atlantoaxial rotatory fixation, the radiologic definition of these conditions is not clear. Patients and Methods. Thirty-three consecutive pediatric patients (average age 8.5 years, range 2–18 years) with acute acquired torticollis were analyzed. All were neurologically intact. Anteroposterior and lateral radiographs were obtained in all atlantoaxial computed tomography scans in 31 patients (dynamic in 23 and static in 8). Twelve age-matched patients with normal cervical spines were also analyzed with dynamic computed tomography as normal controls. Atlantoaxial rotatory subluxation, atlantoaxial angle, center of rotation, and presence of atlantoaxial rotatory fixation were analyzed in each computed tomography. All patients were treated conservatively. Eight had control dynamic computed tomography scans at the end of the treatment. Results. All patients had atlantoaxial rotatory subluxation ≤3 mm. On dynamic computed tomography, the range of atlantoaxial rotation was 30.4° (range 11–54°) toward deformity and 28.3° (range 18–54°) away from deformity (P = 0.333). Atlantoaxial rotatory fixation was not noted in any of the patients. The same measurement for the normal control group was 28° (range 5–41°) (P = 0.770). Of the eight patients with repeat control computed tomography, the atlantoaxial rotatory subluxation was 26° before and 29° after treatment (P = 0.691 to right and P = 0.199 to left). The center of rotation was within dens in 15 of 19 patients, outside dens in 2 of 19, and undetectable in 2 of 19. In the control group, it was within dens in 8 of 11, outside dens in 2 of 11, and undetectable in 1 of 11. All patients were symptom free at the end of the conservative treatment. Conclusion. We could not demonstrate the presence of atlantoaxial rotatory subluxation or atlantoaxial rotatory fixation in our series of 33 consecutive pediatric patients with acute torticollis. Our findings suggest that the existence of these phenomena are doubtful, although not associated with acute acquired torticollis. Acute acquired torticollis is not necessarily the sign of a pathologic condition of the atlantoaxial joint. Finally, it is probably not necessary to obtain computed tomography scans (static or dynamic) in this group of patients at the time of presentation.