In view of her non-union and continued disabling symptoms I would offer revisional surgery. From the CT and MR sequences I don't believe cranial settling to be an issue at the present time. Indeed, in my experience once a C1/2 fusion is successful the chance of C0/C1 involvement seems significantly reduced.
I would favor a posterior approach to take out the wire and granulation
tissue between C1 and C2. C1/2 transarticular screws augmented by a modified Brooks wiring to hold an interspinous C1/2 tricortical iliac crest graft provides me with >95% fusion rate. If the patient cannot tolerate a post-op collar because of rheumatoid TMJ problems I would place her in a halo.
The issue of vertebral artery ectasia must be considered. However, in our
hands the need to abandon a transarticular screw in a rheumatoid patient is rare.
Chevalier JF, Casha S, Bouchard J, Cho RK, Salo P, DuPlessis SJ, Hurlbert RJ: Vertebral artery ectasia and posterior C1/2 transarticular screw fixation: real or perceived limitations? Canadian Journal of Neurological Sciences 31:S11, 2004
Thanks for asking my input.