Confirmatory cut point has limited ability to make accurate classifications in immunogenicity assays.

Competitive inhibition with excess unlabeled drug is used to confirm the presence of antidrug antibodies (ADA) in study samples. We evaluated specific and nonspecific responses from both drug-naive and drug-treated subjects to identify conditions required by the confirmatory assay to make accurate ADA classifications. Nonspecific signal measured in drug-naive samples used to determine assay cut points was uniformly low and close to the screening cut point. Confirmatory assays performed on incurred study samples with nonspecific responses significantly above the level observed during cut point determination resulted in incorrect ADA classifications. Intensity of confirmatory response should be proportional to the screening response and therefore, to ensure accurate ADA classifications, the confirmatory responses cannot be considered as independent but need to be evaluated in relation to the screening responses.