Depending on the particular methods used, the skin biopsy may have only involved an enumeration of intraepidermal nerve fiber density. While condition of the small-fibers may be grossly noted, it's not automatic that further tests for inflammatory infiltrates would have been attempted.

If the nerves show excessive banching and swelling, though, such an etiology might be suspected. But such an autoimmune attack may not be of the "standard" anti-nuclear antibody variety--it may involve specific antibodies to peripheral nerve components (these can be detected by blood tests) or it may even involve antibodies unique to that individual as a result of molecular mimicry (a lot harder to test for).

The Cornell-Weill center has a test called a ganglioside agglutinin test, developed there, which is designed to grossly test for antibody activity to those peripheral nerve components; it's designed to pick up signs of both the known antibodies and those that may be more individual. A positive on this would generally lead to further testing in an attempt to see if there are known antibodies. Keep in mind, though, that many of these "known" antibodies have only been identified in the last 25 years, and there are probably more out there that numbers of people have (e.g., "non-individual") that have yet to be identified/catalogued.