Rheumatol Int (1996) 16:95-100 © Springer-Vertag 1996 A. Dunky • J. Neumiiller • C. Hiibner • G. F. Fischer P. M. Bayer • E. Wagner • D.W.M. Schwartz W. R. Mayr HLA-B27 determination using serological methods. A comparison of enzyme immunoassay and a microlymphocytotoxic test with flow cytometry and a molecular biological assay Received: 25 November 1995 /Accepted: 29 May 1996 Abstract Typing for HLA-B27 is routinely performed in patients with seronegative spondarthritides. Besides the microlymphocytotoxic test (MLCT), other serological techniques have been developed such as enzyme immu- noassays (EIA) using serum or plasma as a source for the determination of soluble HLA-B27 (sHLA-B27) and flow cytometric (FC) methods. The aim of the present study was to check the accuracy and reliability of the EIA for sHLA- B27 in comparison to the MLCT using antibodies against HLA-B27 and cross-reacting speciflcities (CRS), as well as an FC method and a molecular biological method. Any discrepant results should be typed with the MLCT using a complete panel of anti-HLA-class I antibodies, with FC and with a molecular biological technique. The EIA should also be repeated in those patients, using serum and plasma from a new venipuncture. In 81 patients with rheumatic disorders, the EIA and the MLCT using antibodies against HLA-B27 and CRS were performed. Based on the MLCT with a complete panel of anti-HLA-class I antibodies as a standard, discrepant test results were obtained for 9 out of 81 patients with the MLCT using antibodies against HLA- B27 and CRS and with the EIA. The following wrong re- suits occurred: in the MLCT with anti-HLA-B27 and CRS, there were two false-negative results; in the EIA there were four false-negative and one false-positive results; one sam- ple was undeterminable. In comparison with the MLCT, including the complete panel of HLA-class I antibodies, as well as with a molecular biological technique, typing with FC showed a complete concordance. Our investigations A. Dunky • E. Wagner 5th Department of Internal Medicine, Wilhelminenspital, Vienna J. Neumtiller (~) Ludwig Boltzmann Institute for Rheumatology and Balneology, Kurbadstrasse 10, RO. Box 78, A-1107 Vienna-Oberlaa, Austria Fax: 0043-1-681611-234 C. Htibner - R M. Bayer Central Laboratory, Wilhelminenspital, Vienna, Austria G.W. Fischer • D.W.M. Schwartz • W. R. Mayr Clinical Department for Blood Group Serology, University of Vienna, Vienna, Austria demonstrated that for routine typing for HLA-B27 the MLCT cannot be replaced by EIA because of a significant number of mistypings. The MLCT performed only with antibodies against HLA-B27 and CRS may also lead to typ- ing errors. No errors were detected using flow cytometry. If only serological methods can be performed in a labora- tory a combination of flow cytometry and MLCT could therefore enhance the safety of HLA-B27 typing. Key words HLA-B27 • Microlymphocytotoxicity test • Enzyme immunoassay • Flow cytometry • Rheumatology Introduction Although there is disagreement about the value of the deter- mination of HLA-B27 in the diagnosis of ankylosing spon- dylitis (Bechterew's disease) and Reiter's disease [1-4], this investigation is frequently requested by rheumatologists. The prevalence of HLA-B27 in other seronegative spondar- thritides such as reactive arthritis or psoriatic arthritis is cur- rently under discussion [5, 6]. Although the presence of HLA-B27 in these disorders is not of diagnostic value, its influence on the course of the disease has been taken into consideration by some authors [7, 8]. Serological typing of HLA-B27 is routinely performed by using the microlymphocytotoxicity test (MLCT) accord- ing to the NIH standard method [9]. This method is easy to perform and allows the daily typing of a relatively large number of patients. Nevertheless, the interpretation of the test requires some experience in the exclusion of, for exam- ple, false-positive typing results caused by cross-reactions. A significant disadvantage of the MLCT is the subjective microscopic evaluation of the percentage of lymphocytes that are killed or viable. If just the results of HLA-B27 typ- ing are of interest, only a small panel of test sera including positive and negative control sera, and polyclonal or mono- clonal antibodies (MAB) against HLA-B27 and cross-react- ing specificities (CRS) rather than the whole panel of HLA- class I antibodies are usually used for the test.