RESUMEN
BACKGROUND: Paraneoplastic neurological syndromes (PNS) might present as polyneuropathies (PNP). Because PNS are rare and PNP are frequent, it may be difficult to decide in patients with PNP of unclear or presumably idiopathic etiology whether to test for onconeural antibodies, which are highly predictive for PNS. In this regard, this is the largest study investigating the prevalence of onconeural antibodies in patients with PNP, in order to clarify whether such testing should be standard. METHODS: Of 1842 consecutive patients, 283 were suitable and had stored serum samples for screening for onconeural antibodies (anti-Hu, Yo, Ri, CV2/CRMP5, Ma2, and amphiphysin) by ELISA: 159 patients with PNP of - despite laborious work-up - unknown etiology without cancer, 67 with Guillain-Barré syndrome (GBS), 31 with chronic inflammatory demyelinating PNP (CIDP), and 26 with cancer and PNP. RESULTS: None of the 283 screening samples revealed high concentrations for any of the tested antibodies. Thirteen sera (4.6%) showing positive reactivity in the screening ELISA (11 with PNP of unknown etiology without cancer, one with GBS, and one with CIDP) most likely represented increased background activity, as confirmatory assays (immunoblotting and immunohistochemistry) were negative. Furthermore, none of these 13 patients had diagnosed cancer and 10 with sufficient follow-up data did not develop cancer during follow-up. Interestingly, none of the patients with known cancer and PNP was screened positive for any antibody reactivity. CONCLUSIONS: Our data suggest that routine screening for onconeural antibodies in etiologically unclear and in presumably idiopathic (GBS and CIDP) PNP is not mandatory.
