Anti-drug antibody (ADA) assays are important in development of recombinant-protein therapeutic drugs. Appropriate detection and characterization of ADA in treated patients enables a more thorough understanding of the immunogenicity, safety, and efficacy of therapeutic protein products (Egging et al., 2018; Azadeh et al., 2019; U. S. Food and Drug Administration, 2019). Assay interference in various modalities of immunoassays has been of concern to bioanalytical laboratories, and considerable effort has been dedicated to reaching a better understanding of the potential sources of such interference, as well as to the implementation of strategies to mitigate it, resolve it, or prevent it (Tate and Ward, 2004; Crisino et al., 2014; Pierog et al., 2015; Warade, 2017; Ward et al., 2017). This manuscript describes the unexpected and impactful interference we encountered, originating from the interaction between the murine monoclonal antibody (mAb) we used as a surrogate ADA and human anti-mouse antibody (HAMA) found in human serum used to prepare our positive controls (PCs).

Assay PCs play a crucial role in multiple stages of assay development, validation and implementation. They enable a successful definition of assay format, assist in the establishment of assay cut-points and thresholds, and ensure acceptable assay performance during qualification and validation, as well as during sample testing (Egging et al., 2018; Azadeh et al., 2019; U. S. Food and Drug Administration, 2019). The choice of type of molecule to be used as surrogate positive control in clinical ADA assays has been a subject of considerable debate. Surrogate murine anti-ID (MAID) or anti-complementarity-determining region (anti-CDR) mAbs are the most popular types of controls for clinical ADA assays (Yang et al., 2017). Typically, two or more PCs are used to monitor assay performance, as well as to estimate the relative sensitivity and drug tolerance of the assay (Gorovits, 2009; Kelley et al., 2013; Islam et al., 2018; U. S. Food and Drug Administration, 2019). However, these MAIDs are known to experience interference from human serum (HS) since at least 10 % of the general population has been observed to carry some form of animal-derived antibodies, most often from mice (Tate and Ward, 2004). HAMA interference in various types of clinical assays has been well-documented (Kricka, 1999; Tate and Ward, 2004; Dodig, 2009; Lachance and Levesque, 2014; Oldfield, 2014; Warade, 2017; Ward et al., 2017; Islam et al., 2018; Hu and Ho, 2020). In those reports, the interference was likely caused by the non-specific binding of HAMA to a murine-mAb used as capture-reagent in the assay. This manuscript describes the negative impact of endogenous HAMA via a heretofore unrecognized mechanism. The assay in this case was a bridging ELISA, developed to support clinical development of an anti-influenza B (anti-FLUB) humanized mAb therapeutic. We provide evidence that this effect directly originated from HAMA complexation with the MAID mAb used as a surrogate positive control in the assay. This may be, to our knowledge, the first report of such type of interference. The importance of these findings is also discussed.