--if the neuropathy is suspected of originating from an autoimmunity.

It's often mentioned in concert with testing for Guillain Barre Syndrome. The thinking is that Chronic Immune Demyelinating Polyneuropathy is the slower, more insidious form of GBS, and often has a long-term, relapsing-remitting course but otherwise the two clinical entities are very similar. (I have heard them analaogized to different presentations of multiple sclerosis, just of the peripherla nerves.)

The Berger/Pully article in the Useful Sites section here--"Current Concepts in the Diagnosis and Treatment of Peripheral Neuropathies"--talks about it, in that is says elevated CSF protein levels are frequent, though lumbar puncture and nerve biopsy are usually not necessary to achieve a diagnosis:

http://www.dcmsonline.org/jax-medici...uropathies.htm

It's mentioned more prominently as a diagnostic tool--"obtaining both cerebrospinal fluid and a nerve-biopsy specimen is mandatory to make a definitive diagnosis of the disease, according to criteria of the American Academy of Neurology"--by this article in the New England Journal of Medicine (originally published in Germany):

http://171.66.123.143/cgi/content/extract/352/13/1343

It's also discussed in the Merck Manual (Sec. 14, Chapter 183):

http://www.merck.com/pubs/mmanual/se...er183/183f.htm

Much of the referencing here goes back to the following article:

Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology . 1991; 41: 617–618. (I have yet to find it printed in an Internet accessible area on any of the retrieval systems.)

Of course, there have been many articles that dispute the need for lumbar puncture CSF protein findings to make a CIDP dignosis--interestingly, considering where and by whom Alan's neuro was trained, Dr. Latov and his group are prominent among them:

http://www.neurology.org/cgi/content...0/8_suppl_3/S8

http://www.neurology.org/cgi/content.../12_suppl_6/S2

There's also this revision from the above authors from Germany:

http://www.co-neurology.com/pt/re/co...195629!8091!-1

I could go on, but it seems in summation that, at least originally, CSF protein findings were considered an important diagnostic criterion, but, with the expansion of the diagnostic category, it may now be less important, though its presence does imply an autoimmune disorder that, in absence of central nervous system symptoms, would point to CIDP or a related entity and possible IVIg use.

I know personally, from having spoken with Latov, Chin, Moghekar, and a number of other researchers/clinicans, that CSF testing is part of many center's protocols, if for no other reason than to eliminate infectious meningial causes when neuropathic symptoms are present, and to check for
proteins that might point to MS or other central nervous system demyelinating conditions.