Population-based study of diagnostic assays for Borrelia infection: comparison of purified flagella antigen assay (Ideia™, Dako Cytomation) and recombinant antigen assay (Liaison®, DiaSorin)

BACKGROUND Testing for Borrelia-specific IgM and IgG-antibodies are often performed on a variety of poorly defined symptoms, and isolated IgM results are a frequent finding, which results in diagnostic uncertainty and further testing. We wanted to test the hypothesis that Borrelia-specific assays using recombinant antigens perform differently from assays based on purified flagella antigen. METHODS We compared the use of recombinant antigens (LIAISON DiaSorin, Saluggia, Italy) and purified flagella antigen (IDEIA Borrelia, DakoCytomation, Glostrup, Denmark) in the assay for Borrelia-specific IgM and IgG-antibodies. The assays were tested on an unselected population of serum samples submitted from general practice. A total of 357 consecutive samples for analysis of Borrelia IgM and IgG antibodies. Furthermore, we analysed 540 samples for Borrelia-specific IgM or IgG antibodies first by the IDEIA and, if they were positive, the samples were further analysed using the LIAISON assay. To verify the correctness of the patient's serological status, discrepant samples were analysed by line blots (EcoLine, Virotech). RESULTS In the consecutive series of 357 samples, the IgM assays detected 308 negative and 3 positive samples with concordant results. Compared with the line blot, the IDEIA system produced 21 false-positive IgM results, whereas the LIAISON(R) system produced only one false-positive IgM result. The IgG assays showed 1 positive and 328 negative concordant results. The LIAISON system produced 9 true IgG-positive samples that were not detected by the IDEIA system, but the former produced 4 positive IgG results that were negative by line blot. CONCLUSION Diagnostic assays based on flagella antigen seem to show more false-positive IgM and false-negative IgG results than assays based on recombinant antigens. The latter may reduce the number of presumably false-positive IgM results and identify more IgG-positive subjects, but this system also produces more false-positive IgG results.