BACKGROUND: The immunological methods for detecting antibodies to erythropoiesis-stimulating agents (ESAs) differ in assay sensitivity. However, this parameter, routinely determined in clinical assays using a high-affinity non-human polyclonal antibody, gives a one-dimensional assessment of antibody detection. We compare three widely used immunological methods and evaluate the ability of each to detect mature human antibodies and human antibodies characteristic of an early immune response. METHODS: The detection of anti-ESA antibodies was compared between a radioimmunoprecipitation (RIP) assay, an electrochemiluminescence (ECL) bridging enzyme-linked immunosorbent assay and a surface plasmon resonance (SPR)-based immunoassay. All three methods were validated for sensitivity, specificity and precision. Specimens from clinical studies or post market testing were categorized as pure red cell aplasia (PRCA) or non-PRCA and then analyzed in each method. RESULTS: Among the antibody-mediated PRCA samples, which contain high affinity neutralizing antibodies, there was strong correlation between all methods. The results from non-PRCA sample analysis show high correlation between RIP and ECL methods; however, differences between the SPR immunoassay and the ECL and RIP were demonstrated. The samples that scored positive in the SPR immunoassay and negative by RIP and ECL were characterized to be of low antibody concentration, contained a high percentage of rapidly dissociating antibodies, or were antibodies of the IgM isotype. CONCLUSIONS: All three immunological methods are appropriate for detection of antibodies associated with antibody-mediated PRCA. However, the SPR immunoassay platform detected an early, low affinity IgG and IgM antibody response as well as detected and characterized a pathogenic antibody response associated with antibody-mediated PRCA.
BACKGROUND: The immunological methods for detecting antibodies to erythropoiesis-stimulating agents (ESAs) differ in assay sensitivity. However, this parameter, routinely determined in clinical assays using a high-affinity non-human polyclonal antibody, gives a one-dimensional assessment of antibody detection. We compare three widely used immunological methods and evaluate the ability of each to detect mature human antibodies and human antibodies characteristic of an early immune response. METHODS: The detection of anti-ESA antibodies was compared between a radioimmunoprecipitation (RIP) assay, an electrochemiluminescence (ECL) bridging enzyme-linked immunosorbent assay and a surface plasmon resonance (SPR)-based immunoassay. All three methods were validated for sensitivity, specificity and precision. Specimens from clinical studies or post market testing were categorized as pure red cell aplasia (PRCA) or non-PRCA and then analyzed in each method. RESULTS: Among the antibody-mediated PRCA samples, which contain high affinity neutralizing antibodies, there was strong correlation between all methods. The results from non-PRCA sample analysis show high correlation between RIP and ECL methods; however, differences between the SPR immunoassay and the ECL and RIP were demonstrated. The samples that scored positive in the SPR immunoassay and negative by RIP and ECL were characterized to be of low antibody concentration, contained a high percentage of rapidly dissociating antibodies, or were antibodies of the IgM isotype. CONCLUSIONS: All three immunological methods are appropriate for detection of antibodies associated with antibody-mediated PRCA. However, the SPR immunoassay platform detected an early, low affinity IgG and IgM antibody response as well as detected and characterized a pathogenic antibody response associated with antibody-mediated PRCA.