The atlantoaxial distance ratio (AADR): proposal and validation of a novel radiographic parameter for the evaluation vertical atlantoaxial dislocation in a preliminary descriptive study
摘要
Vertical atlantoaxial dislocation (VAAD) is a clinically significant but often overlooked component of atlantoaxial instability, yet a reliable, universally applicable radiographic parameter for its quantification is lacking. This preliminary study proposes and validates a novel metric, the atlantoaxial distance ratio (AADR), for the radiographic assessment of VAAD.
Materials and methodsThe AADR is calculated on lateral cervical radiographs as the ratio of the mean vertical distance from the C1 superior tubercles to the C2 inferior endplate, divided by the anteroposterior diameter of the C2 inferior endplate. A preliminary reference range for AADR was determined in 46 asymptomatic volunteers. Subsequently, a retrospective analysis was conducted on 13 patients with VAAD who underwent single-stage posterior reduction and fusion using C2 lateral mass-isthmus screws between January 2022 and April 2023. Pre- and postoperative AADR, atlanto-dental interval (PADI), spinal canal diameter on MRI (d-MRI), visual analogue scale (VAS) for pain, and modified Japanese Orthopedic Association (JOA) scores were evaluated. A composite reference standard (CRS) was established based on a blinded review by an expert panel. The diagnostic accuracy of AADR was evaluated using receiver operating characteristic (ROC) curve analysis.
ResultsThe mean AADR in the normal cohort was 1.87 ± 0.08 (range: 1.80–2.06). In patients with VAAD, the mean preoperative AADR was significantly lower at 1.78 ± 0.10 (p = 0.0015 vs. normal), and it significantly increased to 2.07 ± 0.12 postoperatively (p < 0.0001), a value not significantly different from the normal range (p > 0.05). Surgical reduction also resulted in significant improvements in PADI (from 7.31 ± 1.32 mm to 15.77 ± 1.24 mm, p < 0.0001), d-MRI (from 6.62 ± 1.26 mm to 15.00 ± 1.22 mm, p < 0.0001), VAS scores (from 4.77 ± 1.01 to 0.77 ± 0.60, p < 0.0001), and JOA scores (from 9.08 ± 3.14 to 12.46 ± 2.47, p < 0.0001). The area under the ROC curve (AUC) for AADR was 0.901, indicating high diagnostic accuracy. The Youden’s J index identified an optimal diagnostic threshold at an AADR value of 1.83 with a sensitivity of 88.9% and a specificity of 73.2%.
ConclusionThis preliminary study suggests that the AADR is a promising and reproducible radiographic parameter for quantifying VAAD. An AADR value below 1.83 may suggest the presence of VAAD; however, this proposed threshold requires validation in larger, multicenter studies. Restoring it to the normal range signifies successful reduction. Correcting VAAD is crucial for achieving effective neural decompression and creating optimal conditions for fusion.