Development, validation and application of ELISAs for pharmacokinetic and HACA assessment of a chimeric anti-CD40 monoclonal antibody in human serum

As part of a Phase I chimeric anti-CD40 monoclonal antibody clinical trial, two enzyme-linked immunosorbent assays (ELISAs) were developed for secondary endpoints: 1) for the pharmacokinetic (PK) monitoring of serum antibody levels and 2) for immunogenic screening of human anti-chimeric antibody (HACA) responses. The ELISA is a well established immunoassay, with clear guidelines for validation when used as a quantitative assay. However, these parameters may not always be relevant for a semi-quantitative assay used to assess whether a sample is positive or negative for a novel marker such as an antibody developed against a therapeutic antibody. We report here the development of a quantitative PK ELISA and a semi-quantitative HACA ELISA, and the different approaches of validation to prove each assay are ‘fit for purpose.’ The parameters of linearity (R 2 > 0.99), accuracy (±30%), lowest level of detection (4 μg/ml), intra-assay (coefficient of variation (CV) < 20%) and inter-assay (CV < 20%) variability were assessed for the quantitative PK assay. For the semi-quantitative HACA assay, parameters of linearity (R 2 > 0.99), lowest level of detection, intra (CV < 10%) and inter-assay (CV < 30%) variability were assessed using a surrogate positive control. The validation outcome showed that each assay was robust, reliable and accurate to meet the requirements of the intended analytical application, that being to 1) quantitatively determine the concentration of antibody in the serum and 2) determine whether a sample is positive or negative for human anti-chimeric antibodies. Each assay has been successfully translated for use in a clinical trial with adequate quality controls and acceptance criteria set for monitoring consistency and performance. ► Details the measurement of specific proteins in complex biological matrices. ► Focuses on approaches taken to overcome common problems with bioassays. ► Demonstrates that validation parameters are dependent on the purpose of the assay.