Object: The atlantoaxial joint is the location most and earliest affected in patients with rheumatoid arthritis (RA). In longstanding disease, ligamentous and osseous destruction can progress and involve all cervical segments. If surgical intervention is necessary, some prefer, to be safe, undertaking fusion to the occiput, whereas others advocate 1-level fusion of C1-2. Sparing the occiput (Oc)-C1 segment would allow retention of a considerable amount of physiological range of motion and seems beneficial against subaxial overload. Previous clinical studies on this topic have provided only nonspecific data after short-term follow-up, rendering a segment-sparing approach questionable. The purpose of the present investigation was to assess long-term progression of inflammatory or degenerative destruction in the Oc-C1 segment after isolated C1-2 fusion for RA.
Methods: In a series of 113 consecutive patients with RA-related destruction restricted to the craniocervical junction, 14 individuals underwent Oc-C2 fusion and 99 underwent surgery exclusively at the C1-2 level. After a mean follow-up period of 9.4 years (range 4.9-14.7 years), 46 patients were available for clinical and radiographic examination, including CT imaging.
Results: None of the 46 patients needed additional surgery to extend the fusion to the occiput. Despite marked deterioration in the subaxial cervical spine, in general there were little or no changes in the atlantooccipital region. All but one patient presented with bony fusion of the fixed C1-2 level at follow-up.
Conclusions: The results of this investigation suggest that if the Oc-C1 joint is free of osseous destructions on conventional radiographs and free of abnormalities on MRI scans at the time of surgery (for transarticular fixation and fusion of C1-2), there is a very low risk for relevant destruction in the following 5-14 years. Thus, no prophylactic oligosegmental approach, but rather a segment-sparing monosegmental approach, is preferred, even in patients with high inflammatory levels.