(M1230-06-36) Method Development and Validation of a Sensitive and Robust Human Serum PK Immunoassay for a Novel Bispecific Antibody by an Automated Microfluidic System

Purpose: Antibody-based therapeutics have been evolving from monoclonal antibodies (mAb) to more complex structures containing more than one functional domain to achieve advanced therapeutic effects. These multiple domain therapeutics require unique approaches to bioanalysis in order to characterize their pharmacokinetic properties. BI X is a novel tetravalent bispecific antibody under development for cancer treatment. It is critical to develop a sensitive immunoassay in understanding its clinical pharmacokinetics. The Gyrolab® microfluidic system has been widely used due to its reduced manual labor and enhanced workflow efficiency, while providing high assay sensitivity with large dynamic assay range. The purpose of this task was to develop and validate a sensitive immunoassay using the Gyrolab® platform in supporting clinical pharmacokinetic studies of BI X.

Methods: Reference standard BI X and critical reagents were all prepared in house. The Gyrolab® xP workstation was used. During the method development, the assay conditions were optimized based on selectivity and Gyrolab® Binding Profile Images to obtain required sensitivity and dynamic range. The conditions for optimization included Gyrolab® CD selection, assay format, minimum required dilution (MRD), anti-idiotypic antibody pairing for capture and detection, degree of labeling for biotin- and Alexa Fluor 647-labels, assay buffer, and needle wash buffer and wash steps. In the optimized method, Gyros’ Bioaffy® 1000 CDs containing streptavidin beads were used. BI X was captured on the CD with a biotinylated anti-idiotypic mAb against the target A domain and detected with an Alexa Fluor 647 labeled anti-idiotypic mAb against the target B domain. Calibration standard, QC and samples were diluted 10 times by adding 10 μL sample in 90 μL of Rexxip® H-X buffer. Additions of capture reagent, analyte, detection reagent, and wash solutions to the microstructures on the CD were fully automated processes. Following a wash to remove unbound Alexa Fluor 647, the fluorescence intensity was measured, which is proportional to the quantity of BI X in the assay range. The assay validation was carried out on the following parameters: calibration standards and curve range, accuracy and precision, dilution linearity and hook effect, selectivity, carryover, and analyte stability.

Results: The calibration standard curve range was validated from 10.0 - 2000 ng/mL for BI X in human serum with MRD at 10. Accuracy and precision of the method were determined by analyzing QC samples at five different concentrations within the calibration standard curve range, demonstrated %bias from -14.7 to -0.1% and percent coefficient of variation (%CV) from 1.2 to 4.2% for intra-run, %bias from -8.5 to 5.0% and %CV from 5.7 to 9.8% for inter-runs, and total error from 6.5 to 14.2%, respectively. QC Intra-batch Precision (%CV) was 1.2% to 4.1%. Matrix selectivity in normal, hemolyzed, lipemic and colorectal human serum samples was evaluated in unspiked blank and in blank spiked with BI X at 10.0 and 2000 ng/mL and shown that at least 80% of the tested lots had a recovery between 75.0-125% when tested at the LLOQ, at least 80% of the tested lots had recovery between 80-120% at the ULOQ, and at least 80% of the unspiked lots tested were BQL. Dilution linearity was tested in human serum samples prepared at 200,000 ng/mL and diluted with human serum using a series of dilution factors and demonstrated the maximum in range dilution of 20000-fold. No hook effect was observed in the dilution linearity test. Analyte-specific carryover was tested to demonstrate that the needle wash step in the method was effective in preventing analyte-specific carryover between sample pipetting. Pooled human serum was pipetted with the same needle before and after pipetting a carryover test sample of 2000 ng/mL and all the testing samples showed BQL before and after high concentration carryover sample pipetting. BI X stability was evaluated by QC samples prepared in human serum and found stable under the following stability conditions: five freeze/thaw cycles in freezer storage set at -20°C and -70°C, 24 hours at Room Temperature, 19.5 hours at 4°C and 6 hours at room temperature for Processed Stability, and Long-term Storage Stability 36 months at -20°C and -70°C.

Conclusion: A human serum PK immunoassay has been developed and validated successfully for BI X on the Gyrolab® platform. The validation results demonstrate that the assay is sensitive and robust to support BI X clinical pharmacokinetic studies.

Acknowledgements: The authors would like to thank Christine Grimaldi, C. Elisa Oquendo, Henry Zhao, Chuck Ye, Steven Anderlot, Jolaine Savoie, Andrey Konovalov, Hamid Samareh Afsari, and QPS for their support of this project.