(T0930-02-08) Validation of an Anti-Drug Antibody Assay Against an Oncolytic Virus Immunotherapeutic

Purpose: Oncolytic viruses (OV) immunotherapy, an emerging class of cancer therapeutic agents, involves the use native or genetically modified viruses that selectively replicate in tumor cells and induce systemic anti-tumor immunity¹. The OV under study is a novel genetically modified adenovirus serotype-5 (Ad5) that selectively targets tumors. Such biotherapeutics can elicit an antibody response that can affect product safety and efficacy. Immunogenicity assessments during clinical trials that can characterize these immune responses are crucial for these biotherapeutic development programs². However, the presence of pre-existing antibodies to adenoviruses in the human population may cross-react with the OV capsid proteins and mitigate Ad5-based therapeutics³. This study was undertaken to measure the treatment-emergent anti-drug antibody (ADA) response in a patient population who may already contain pre-existing antibodies to Ad5.

Methods: The presence of pre-existing antibodies prevents the use of a traditional multi-tiered immunogenicity assessment, consisting of a screening, confirmatory and titer assays. In this study, only screening and titer assay formats were validated using an electrochemiluminescence platform (Figure 1). Since these pre-existing antibodies cross-react with the capsid proteins of the OV, a cut point estimation from screening individual normal human serum was not feasible. Therefore, a plate-specific cut point was established in each assay plate using specialty serum (NC)⁴. To establish positive controls (PCs) for the method, commercially available anti-adenovirus Type 5 polyclonal antibody was spiked in NC to prepare the low-concentration positive control (LPC), the medium concentration positive control (MPC), and the high-concentration positive control (HPC) at 100, 2000, and 10000 ng/mL, respectively. Validation of the ADA method involved two phases, Validation Phase I and Validation Phase II. Validation Phase I was performed by two analysts and included the establishment of the HPC titer and the Minimum Significance Ratio (MSR) over a minimum of ten runs. The MSR reflects the smallest fold-change in the titer values that can be considered statistically significant (P < 0.05). Validation Phase II involved evaluating sensitivity, precision, drug tolerance, titration assay linearity, specificity, selectivity/interference, robustness, prozone effects, and stability of the method.

Results: To calculate a plate-specific cut point, a multiplication factor of 1.25 was introduced during data analysis using the equation⁴: Cut point = Mean NC signal of each plate × 1.25. The dilution factor of the sample that intersects the model at the cut point was reported as the titer. The MSR was calculated according to USP 1106⁵ using a minimum of fifty (50) data points and the following formula: MSR = 10^t-√2-SD, where SD is the overall standard deviation of the titers in log scale, and t is the threshold from Student’s t-distribution with n − 1 degrees of freedom. Data from Validation Phase I determined the HPC titer and the MSR to be 118324 and 2.39, respectively and data from Validation Phase II demonstrated the method to be precise, selective, sensitive to 4.38 ng/mL (Figure 2), drug tolerant, specific, linear, robust, and stable (Table 1). Following the successful validation of the ADA assays, human clinical samples were analyzed for the presence of treatment-emergent ADA. In the screening assay, a clinical sample was considered positive for ADA if it generates a mean signal greater than the cut-point. Samples deemed to contain ADA were then analyzed in the titer assay to determine the titer of the sample. Any two samples that have a fold-difference greater than the MSR determined during the method validation study was considered significantly different from each other.

Conclusion: In conclusion, we report here the successful validation of an ADA assay in the presence of pre-existing antibodies using a non-traditional approach. The results obtained for each validation parameter demonstrated that the method was suitable for measuring the titer of ADA response against the Ad5 OV in human clinical samples.

References: 1. Bray, F., Laversanne, M., Weiderpass, E., & Soerjomataram, I. (2021). The ever-increasing importance of cancer as a leading cause of premature death worldwide. Cancer, 127(16), 3029–3030.
2. Wadhwa, M., Knezevic, I., Kang, H. N., & Thorpe, R. (2015). Immunogenicity assessment of biotherapeutic products: An overview of assays and their utility. In Biologicals (Vol. 43, Issue 5, pp. 298–306). Academic Press.
3. Fausther-Bovendo, H., & Kobinger, G. P. (2014). Pre-existing immunity against Ad vectors: Humoral, cellular, and innate response, what’s important? In Human Vaccines and Immunotherapeutics (Vol. 10, Issue 10, pp. 2875–2884). Landes Bioscience.
4. Collet-Brose, J., Couble, P. J., Deehan, M. R., Nelson, R. J., Ferlin, W. G., & Lory, S. (2016). Evaluation of Multiple Immunoassay Technology Platforms to Select the Anti- Drug Antibody Assay Exhibiting the Most Appropriate Drug and Target Tolerance. Journal of Immunology Research, 2016.
5. United States Pharmacopeia (2022). General Chapter, 〈1106〉 Immunogenicity Assays— Design and Validation of Immunoassays to Detect Anti-Drug Antibodies. USP-NF. Rockville, MD: United States Pharmacopeia.

Acknowledgements: The authors declare no conflicts of interest.

Figure 1. Schematic representation of an immunogenicity assay to measure the anti-drug antibody response to an oncolytic virus.

Figure 2. Assay Sensitivity. Graph of assay Log10(signal) vs Log10([PC] ng/mL) across 4 runs. Sensitivity was determined by back calculating the concentration at which the sensitivity curve meets the cut point. The mean value of sensitivity between runs (n=4) was 4.38 ng/mL.

Table 1. Summary of Validation Criteria