Radiographic measurements of cervical alignment, fusion and subsidence after ACDF surgery and their impact on clinical outcome

Some recent studies indicate correlations between cervical alignment and clinical outcome after anterior cervical discectomy and fusion (ACDF) surgery. However, there still are no established criteria for the interpretation of alignment, fusion and subsidence in relation to clinical outcome.

A retrospective analysis of 208 radiographs of patients following ACDF with stand-alone PEEK cage implantation was performed. The measurements were obtained on plain radiographs in lateral and anteroposterior projections as well as flexion/extension radiographs. Cervical alignment was measured using the Gore, Laing and Cobb methods; fusion was evaluated by an assortment of radiographic hallmarks: the presence of bridging bone, the Cobb angle and the distances between the tips and bases of the spinous processes of the operated segments, respectively. For assessment of subsidence, we used the Mochida method in addition to ventral and dorsal segmental height reduction. Correlation analysis between the different radiological characteristics and clinical outcome at a minimum follow-up of 12 months was conducted.

Two hundred and eight patients were evaluated for alignment, fusion and subsidence. Cervical alignment using the Gore and Cobb methods correlated among each other, but failed to exhibit significant correlation with clinical outcome. Interpretation of fusion rates varied greatly (43.9 to 89.4%) depending on the criteria used. Pearson coefficients between radiographic presence of pseudarthrosis and the measurements of the spinous process distances (0.595; p < 0.001), the Cobb angles (0.187; p = 0.007) and the presence of bridging bone (0.224; p < 0.001) each exhibited statistical significance. None of the methods employed significantly correlated with clinical outcome. Regarding subsidence, we found rates of 62%, 48% and 27% using the Mochida, ventral and dorsal segmental height reduction assessment methods, respectively. Pearson correlations between pairs of Mochida/ventral (r = 0.39; p = 0.66) and Mochida/dorsal (r = 0.007; p = 0.921) height reduction assessment methods were weak and no significant correlation between subsidence rates and clinical outcome was shown.

All measured parameters varied depending in the measurement method used. This was most pronounced for fusion. There was a moderate positive correlation between neck pain and subsidence as measured by the Mochida method.