Correlation of screening and confirmatory results in tiered immunogenicity testing by solution-phase bridging assays

Biotherapeutic proteins induce undesired immune responses that can affect drug efficacy and safety. For this reason, immunogenicity assessment is an integral part of drug development and is mandated by the regulatory authorities. Immunogenicity is typically evaluated by a tiered approach consisting of a screening assay followed by a competitive inhibition with unlabeled drug serving as confirmatory assay and additional characterization of the immune response. The confirmatory assay is intended to reduce the number of false positive responses generated in the screening tier and ensure that all samples are correctly classified as positive or negative. The positive-negative sample decisions are based on screening and confirmatory assay cut points that are statistically derived through evaluation of drug-naive samples. In this paper, we describe the analysis of cut point data for the presence of statistical correlation between the screening and confirmatory results. Data were obtained from validations of solution-phase bridging assays for detection of anti-drug antibodies against monoclonal antibody therapeutics. All data sets showed moderate to strong positive correlation, indicating that the screening and confirmatory assays were not independent and were likely to generate similar information. We present theoretical evidence that correlated results may be a general feature of the tiered approach when the same test platform is used for both screening and confirmatory assays. The competitive inhibition test, therefore, may be of limited value beyond reduction of the overall false positive rate. Our results indicate that similar sample results could be obtained by using just the screening assay with the false positive rate set to 1%.