Title: Subject Pretreatment vs. Time Point Treatment Variation in ADA Assays: Validity of Normalizing Results to Subject Pretreatment Signals

Abstract: The format for many immunogenicity assays is based on the values of negative or non-neutralizing samples and then inferring, by some pre-determined shift in the negative value, whether a sample is positive or remains negative. While the design of most immunogenicity assays is such that negative samples should not register or give a signal, the reality is often individual samples do give some signal in assays. This is especially true when the assay is a cell-based functional assay designed to measure neutralizing potential of an antibody response. The variation in individual samples in cell-based assays is often due to additional factors (ex. growth factors, cytokines) that are present in samples at different levels and that may affect the biological cascade of the assay in ways outside of the designed drug relevant pathway of the assay. In some cases, with high levels of variability in negative samples, a baseline pre-treatment sample from an individual may be used as a benchmark to assess future responses. This methodology of assessing post-treatment responses against pre-treatment responses may be valid if the variability between individuals is responsible for all sample variation. However, it is also important to verify if individuals not only differ from each other, but also, if samples from the same individual over different time points also varies. We have taken advantage of pre-treatment screening visits to collect multiple samples pre-treatment from the same individuals and compare variation across individuals versus variation across days. In the data we will present, there was much higher variability across days than across individuals, making the possibility of using pre-treatment response baselines an invalid methodology to assess neutralizing potential of anti-drug antibodies in post-treatment samples.