Evaluation of methods to detect and characterize antibodies against recombinant human erythropoietin

Background: From May 1998 to November 2000, 13 European patients developed antibody-mediated pure red cell aplasia during treatment with recombinant human erythropoietin (rHuEPO), reinforcing the need for analytical testing for antibodies against erythropoietic agents. Specimens from 8 patients were provided for further antibody testing and characterization. Methods: We evaluated 4 analytical methods with these sera: radioimmune precipitation (RIP), enzyme-linked immunosorbent assay (ELISA), biosensor immunoassay, and a bioassay for identification of neutralizing antibodies. Results: The RIP, biosensor immunoassay, and biological assay performed equally to detect and quantify anti-rHuEPO antibodies. The ELISA failed to detect antibodies in 2 patient samples. The biosensor immunoassay could determine antibody isotypes, subclasses, and dissociation rates and the bioassay could determine whether these antibodies were able to neutralize the biological effect of the drug; the other assays could not make this determination. Conclusion: We recommend the use of the biosensor immunoassay followed by a bioassay to best identify and characterize anti-rHuEPO antibodies. The biosensor immunoassay allows for identification of all classes and subclasses of immunoglobulins and is a preferred method for the identification of lower-affinity antibodies. The bioassay is needed to determine if the antibodies identified have the capacity to neutralize a biological effect of the drug in a cell-based system.