Serodiagnosis of neuroborreliosis: comparison of reliability of three confirmatory assays.

The sensitivity and specificity of three confirmatory assays for the serodiagnosis of neuroborreliosis were investigated. Samples from 96 patients with proven neuroborreliosis, 80 healthy volunteers, 20 patients with neurosyphilis and 20 patients with recent infections with Epstein-Barr virus (EBV) were tested for borrelial antibodies by immunoblotting, Borrelia burgdorferi sensu lato sonicate EIA following pre-absorption of cross-reactive antibodies (Abs-EIA) and by a so-called RECO-EIA using the following recombinant borrelial proteins as antigens: a 14 kDa-internal flagellin fragment, the outer surface protein C (23 kDa) and the high molecular mass protein p83 (83 kDa). The immunoblots were evaluated according to the criteria published by Engström et al. and Hauser et al. An evaluation of IgM and/or IgG antibodies revealed a considerably higher sensitivity for the RECO-EIA (94%) compared to the Abs-EIA (82%, P < 0.0001). Evaluation of the immunoblot according to the criteria of Hauser was significantly more sensitive than according to the criteria of Engström (89 vs 51%, P = 0.0003). A higher sensitivity was demonstrated for IgM (54 vs 22%) and IgG antibodies (64 vs 24%). When both findings from RECO-EIA and immunoblotting were considered, positive findings in the first step assay (sonicate EIA without pre-absorption) were confirmed in 97% of patients. When samples were tested for IgM antibodies, the specificities of the three confirmatory assays did not differ significantly, but in the case of IgG antibodies, the immunoblot (Hauser: P = 0.013; Engström: P = 0.004) and the RECO-EIA (P = 0.02) were more specific than the Abs-EIA. It is concluded that the immunoblot (evaluated according to Hauser) and the RECO-EIA are both suitable as confirmatory assays in the serological diagnosis of neuroborreliosis. Monoclonal antibodies are mandatory tools in the evaluation of the immunoblot.