A Data-driven Stepwise Approach to Cross Validation of Clinical Anti-Drug Antibody (ADA) Assays

Cross-validating bioanalytical assays between laboratories is challenging, and antidrug antibody (ADA) assays are particularly difficult to cross-validate due to the lack of regulatory guidance specific to this topic. As a result, different companies have adopted varying approaches. Currently, bioanalytical practices for ADA assay cross-validation are not harmonized. Therefore, it is important to share individual company experiences and practices to facilitate harmonization for industry standards. This manuscript presents Boehringer Ingelheim's strategy and practice, referred to as a "data-driven stepwise approach," for clinical ADA assay cross-validation. In Step 1, the ADA assay is first validated at the reference laboratory and then qualified at the comparator laboratory. In Step 2, both laboratories demonstrate comparable responses through cut-point evaluation, confirmed by statistical analysis. In Step 3, identical blinded test samples are analyzed at both laboratories to confirm consistent classification of results as ADA-negative or ADA-positive. Upon satisfactory completion of these steps, the ADA assay is considered successfully cross-validated from the reference laboratory to the comparator laboratory. Case studies are also presented.