Methodology

Strength in numbers – The challenges of a robust reference value

Core hypothesis: High signal = high amount of ADA present in the sample

One measurement of signal intensity could potentially replace a multi-tier sequence of assays which provide only positive “+” / negative “-“ information.

To realize a reliable ADA analysis in a single tier approach, a set of prerequisites are required:

Appropriate increased cut-points to minimize false-positive samples:

Calculate positive/negative threshold („cut-point“) with decreased „False Positive Rate“ (e.g. 1% instead of 5% „FPR“): Thereby, confirmatory tier  assay is replaced by signal strength: positive sample status is confirmed due to high-level signals.

Minimize potential systematic bias caused by: ASSAY-TO-ASSAY VARIATION

Absolute signal intensity varies from assay to assay – Ensure and confirm during assay development that normalization (“S/N” = sample signals divided by negative control samples)  is sufficient to compensate assay-to-assay variation of measured signals

Minimize potential systematic bias caused by: DRUG INTERFERENCE

The presence of drug will decrease signals – this interferes with signal strength evaluation: Ensure and confirm improved drug tolerance during assay development and validation: While in current multi-tier approach a qualitative drug tolerance is sufficient (as long as signals are still positive decrease doesn‘t matter) one tier analysis requires quantitative drug tolerance

Minimize potential systematic bias caused by: HOOK EFFECT

A prozone effect / oversaturation with ADA will decrease signals and thus also interferes with signal strength evaluation: Ensure and confirm absence of prozone effect in the relevant concentration range (ng/mL – mg/mL) during assay development and validation; if prozone effect is there, include titration controls

Chimera has established and optimized an adapted protocol for electrochemiluminescence (MSD) detection of ADA presence in a one-step bridging assay set-up. The combination of Chimera’s AnySource® strategies and the reliable MSD technology enables robust performance, high drug tolerance, and a broad dynamic detection range.

The presence of ADA was first evaluated in a validated conventional 3-tier approach. Normalized signals (S/N) were recorded in screening assay tier for comparison with screening cut-point, screening positive samples were analyzed for signal inhibition (%SI) in confirmatory tier and confirmatory positive samples were serially diluted for titer determination, respectively.

Additionally, the comparison of the numerical value of S/N  with %SI and titer enabled determination of Pearson correlation coefficient for these parameters. Positive/negative status of S/N values were also determined with an adapted 1% FPR cut-point. The amount of samples with the same positive/negative status for 1% FPR S/N analysis vs. confirmatory status was calculated in % of total screening-positive samples.

Conclusions

There is more than one way – Confirmed correlation is not a status, but a process

Our findings indicate that S/N analysis could support a 2-tier strategy combination of confirmatory assay and S/N evaluation. This replaces the especially elaborate titer determination (requiring multiple additional dilution samples and consuming high amounts of biological matrix). High robustness against drug and prozone interference is mandatory for this approach and should be confirmed accordingly in selection of assay set-up and validation parameters.

The observed general feasibility of a 2-tiered ADA testing strategy thereby enables more efficient sample evaluation but demonstrates also the current limitations of S/N analysis, especially in relation to the confirmatory tier.

Currently, a high correlation between %SI and S/N values is not necessarily equivalent to fully identical results for “confirmatory positive” and “S/N positive” status – and vice versa. Thus, a systematic evaluation of an increased dataset is recommended for an improved understanding of the underlying drug/ADA interactions. This is now in progress as chimera continues to subject all available ADA data to the innovative S/N analysis strategy.

Until it is fully clarified which effect causes the observed systematic pattern of low %SI vs. high S/N response, confirmatory assay %SI and “one-tier” S/N should be recorded in parallel for further in-depth analysis of signal strength information.

Results do, however, vary in impact from study to study. The evaluated studies reveal both shared correlation patterns and study-specific effects. A case-to-case evaluation of immunogenicity risk assessment and applicability of the 2-tier or even a potential 1-tier strategy should be applied during method development and validation for each study.

S/N information does provide valuable additional information for immunogenicity analysis and should thus be an important part of ADA assay set-up in the future. With expanded experience in S/N analysis, this could open a reliable pathway to the aspired streamlined 1-tier immunogenicity set-up.

The inherent potential of the 1- and 2-tier strategy to expedite data delivery and reduce resource needs in clinical development by enhanced operational efficiency clearly justifies an extended in-depth testing of this innovative approach, including further validation studies and regulatory engagement to optimize implementation in clinical trials.