(T1530-02-09) False Positive or Not? – Assessing Potential Interference of Soluble Target to ADA Detection for a Clinical Stage Program

Purpose: Electrochemiluminescence bridging assay is the most used format to detect anti-drug antibodies (ADA) in drug development and clinical studies for biologics. A challenge often encountered with this assay format is target interference, where a soluble target can bridge the capture and detection reagents and generate a false positive signal. During ADA assay development, target interference is often assessed. However, it can be challenging to assess the likelihood of target-induced ADA positivity in actual patient samples. This is due to several reasons. Firstly, the target level reflecting the relevant study population and treatment may not be available or may be unknown. Secondly, the physiologically relevant form of the target may not be known or available for experimental testing, particularly if the target exists in multiple forms. Therefore, it is critical to assess the true impact of target on the ADA assay or ADA data during the clinical stage. Here, a systematic approach was taken to confirm the validity of ADA positive results reported in a Phase 1 study for biologic-x.

Methods: Biologic-x is a humanized monoclonal antibody against an immune check point protein and is currently in Phase 2 clinical development for solid tumor treatment. High ADA incidence (ranging from 50% to 70%) was detected in Phase 1 and Phase 2 studies, which was unusual for the class of biologics targeting the same target and in largely immune compromised cancer patients. While there is currently no direct evidence linking the high incidence of ADA of biologic-x to any specific clinical observations, a therapeutic that has a substantially higher rate of ADA may be less favored compared to other drugs in the same category. To assess whether the detected high ADA incidence was due to soluble target, a ligand competition assay was established. A large panel of commercial and in-house antibodies was screened to identify competitive antibodies. Then, a panel of study samples, including ADA negative and ADA positive with various titer levels, was processed and tested in parallel in the screening and confirmatory tiers with and without the competitive antibodies to assess potential impact of soluble target to the ADA detection.

Results: First, several competitive antibodies were identified which showed near complete inhibition of biologic-x binding to the target in the in vitro competitive ligand binding assay. All identified competitive antibodies were shown to inhibit the soluble target induced signal in the soluble target spiked samples. Applying two of the identified competitive antibodies to selected study samples, the data demonstrated minimum change in ADA assay signals in presence of the competitive antibodies across tested study samples.

Conclusion: The results suggested that soluble target present in the study samples was not causing false ADA positive results and confirmed that the high ADA incidence was specific to biologic-x. This case highlights a common challenge in clinical ADA assessment and presents strategies that can be beneficial for the field of clinical immunogenicity assessment.

Acknowledgements: All authors claim no conflict of interest.