I think the doc is looking at the o-bands, however the presence in serum as well as CSF is where he's thinking it's not MS. A study suggested that o-bands in the serum is more indicative of neuropathic disorders over MS. Although that study did also find one pt who had both, out of 146.

Return to stats class for a moment, and view statistical significance... Starting pop. 1874, positive pop. 146, variance found 1. Don't even have to do the math to see that it is a insufficient test regulating data to correlational study and requires much more testing. You can find the excerpt here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1014760/

Now some of it could also be the location of the white matter. Different areas in the brain point more to one disorder over another. Unfortunately, MS doesn't hold the copywrites for demyelination. For MS they're largely looking for lesions in and around the corpus callosum. So that might be another reason the doctor is trying to rule it out. Especially since migraines can cause an increase in lesion-like showings on an MRI.

However given the doctor's argument that there are not lesions in the spine... I would definitely go for a second opinion. And given the presence of o-bands both in the serum and CSF, both of which are indicative of demyelination linked to probably both a CNS disorder of unclear dx at this time, as well as a PNS disorder.

Even if they are disorders outside their specialty, they should at least assure you there is clearly something medical going on, and even if they don't think MS they should be able to give you some ideas on what it could be then.