(T1130-06-36) Method Development of BI-X ADC Gyrolab Assay for the Determination of Anti-BI-X Antibodies

Purpose: BI-X is an antibody drug conjugate (ADC) which consists of IgG1 antibody and highly potent anthracycline toxin by site specific SortaseA enzymatic conjugation. ADC drugs have the potential to elicit the formation of anti-drug antibodies (ADA), which have the capability to impact safety and efficacy and to alter pharmacokinetics. Immunogenicity analysis is required to assess the presence of ADA. The aim of this work was to develop a Gyrolab ADA assay for BI-X in human serum to minimize matrix interference that was observed in the BI-X ECL ADA assay.

Methods: To develop a suitable ADA assay for BI-X ADC program, both Electrochemiluminescence (ECL) and Gyrolab immunoassay platforms were explored for assay development. The bridging format with acid dissociation was used in ECL and Gyrolab platforms. The Gyros ADA method utilized offline acid treatment and neutralization with 6 µg/mL of Biotin and Alexa Fluor 647 (AF647) conjugated ADC and 1.5M Tris, pH 10 at optimal condition. Biotin and AF647 labeled ADC and ADA formed complexes that bound to streptavidin coated particle in Bioaffy CD 200. In bridging ECL assay, the complexes of biotin and sulfo-tag labeled ADC with ADA bound to streptavidin coated MSD plate. The cut points of 33 normal individual sera, selectivity and drug tolerance were evaluated in both ECL and Gyrolab platforms.

Results: Both platforms showed adequate levels of sensitivity and drug tolerance. However, high variability in normal human individual sera was observed in screening and confirmatory tiers in ECL assay, but not in Gyrolab ADA assay (Figure 1). The matrix selectivity was evaluated in ECL and Gyrolab platform. The recovery of selectivity in ECL ADA assay did not meet acceptance criteria (75%-125%); however it passed criteria in Gryolab ADA assay (Figure 2). The variability in individual sera may be caused by matrix interference. Short column flow-through time on Gyrolab platform minimized the matrix effect.

Conclusion: The matrix interference, in the ECL ADA assay, increased variability in individual sera and the rate of outliers in the cut point assessment which resulted in a high cut point. The matrix interference was minimized in Gyrolab platform.

Acknowledgements: Thanks Hamid Samareh_Afsari, Stephanie Kostuk, and Steve Anderlot for reagents support.

Figure 1. Box plots of S/N and % inhibition of 33 naïve human individual sera in ECL and Gyrolab. The S/N of each individual in screening assay is shown at left and %inhibition of each individual in confirmatory assay is shown at right.

Figure 2. Recovery of selectivity of individual matrix in ECL and Gyrolab. Selectivity recovery values were calculated based on analysis of individual sample-generated responses at a given PC concentration vs. signal generated by a PC spiked normal serum pool. Two dash lines indicate 75% and 125% recovery, respectively.