[Imaging study on determining the rationality of atlantoaxial fixation angle based on the ratio of line segments between anatomical markers on lateral X-ray films]

Objective: To propose a method to determine the unreasonableness of the fixed angle in posterior atlantoaxial fusion surgery based on the ratio of line segments between anatomical landmarks of the atlantoaxial joint. Methods: A cross-sectional study was conducted. According to the inclusion criteria, a screening was performed on the database of asymptomatic volunteers who had full-spine lateral X-ray films taken at the Second Affiliated Hospital of Wenzhou Medical University from May 2016 to May 2021. A total of 207 volunteers were included, comprising 98 males with an age of (40.68±13.87) years and 109 females with an age of (42.64±14.45) years. On the lateral X-ray film, a line (L) parallel to the posterior margin of the odontoid process was drawn at the posterior edge of the lower articular surface of the axis (a), intersecting the atlas at points b, c, and d. The line segments ab, bd, bc, and the C1-C2 angle were measured, and the ratios of bd/ab and bc/ab were calculated. The ability of bd/ab and bc/ab to predict the unreasonable fixed angle of the atlantoaxial joint (≥22°) was analyzed by receiver operating characteristic (ROC) curve analysis in both male and female. The areas under the ROC curves (AUC) were calculated, and the performance of the two prediction methods was compared using the Delong's test. The cutoff value for distinguishing the unreasonableness of the C1-C2 angle and the sensitivity and specificity were calculated. Results: The ROC curve analysis in the male group showed that the AUC of bc/ab for predicting the unreasonable C1-C2 angle was 0.791 (95%CI: 0.696-0.867, P<0.001), with a cutoff value of 0.449, sensitivity of 97.3%, and specificity of 70.0%. The performance was significantly better than that of bd/ab (cutoff value 1.100, AUC=0.532, 95%CI: 0.428-0.634, sensitivity 26.3%, specificity 83.3%, P<0.001). The ROC curve analysis in the female group showed that the AUC of bc/ab for predicting the unreasonable C1-C2 angle was 0.804 (95%CI: 0.745-0.852, P<0.001), with a cutoff value of 0.488, sensitivity of 90.5%, and specificity of 58.6%. The performance was significantly better than that of bd/ab (cutoff value 0.960, AUC=0.687, 95%CI: 0.624-0.748, sensitivity 90.5%, specificity 44.8%, P=0.041). Conclusions: The bc/ab value can be used as an effective indicator to predict the unreasonable C1-C2 angle in posterior atlantoaxial fusion surgery with high diagnostic accuracy. The cutoff value for males is<0.449, and for females is<0.488.