by Dr. Michelle Miller, Bioanalytical Consultant
After a busy spring and summer season of bioanalytical conferences, we are gearing up for a full schedule this fall. From EIP in Portugal to WRIB in Texas and SSF in Ohio, Celerion’s scientific leaders have spent countless hours engaged with others in the bioanalytical community this year and have found a now-familiar topic being discussed at every event: Should Signal-to-Noise replace Titer in immunogenicity testing? Here, we establish the context and provide a few of the key considerations for understanding the nuances of this relevant and important discussion.
Why do we test for immunogenicity using a 3-tier approach in the first place?
In the early 2000’s, several reports of pure red-cell aplasia (PRCA) after treatment of patients with human recombinant erythropoietin started a series of investigations that eventually led to a new concern for unwanted immune responses, specifically the formation of antibodies against dosed therapeutics in clinical trials. Although the risk of immunogenicity following therapeutic administration is known to be complex and multifaceted, strategies to detect and characterize these unwanted immune responses became a critical component of safety testing strategies in subsequent years.
Early adoption of anti-drug antibody (ADA) testing for therapeutics lacked clear definition as the first official guidance documents from global health authorities were not released until the mid to late 2000’s. As such, drug developers first drew from experience monitoring antibody responses in the vaccine field where titer was the critical measurement of efficacy. It is important to note that titers for vaccine programs are magnitudes higher than titers for most biotherapeutic drug responses. As such, scientists in the industry sought to adapt the testing strategy to better fit biotherapeutics, leading to what we now know as the three-tier approach.
Titer was still considered the main read-out of ADA positivity, but because this method is low throughput and resource intense, titering every sample from every patient was a daunting task. Therefore, a tiered system was designed. This starts with a screening assay where samples are run at a single concentration and any sample with ADAs detectable above background (signals statistically higher than known ADA-negative samples) is identified as positive. The statistical threshold for the screening tier is set to include a 5% “false positive rate” or FPR to ensure that no low positives are missed. However, there was concern that the screening tier could detect non-specific signals for some samples. To address this, a second assay, the “confirmatory” tier, was added that identifies non-specific positives and uses a 1% FPR, thus identifying what are considered “true” ADA positive samples. Finally, samples that both screen and confirm positive are evaluated in a third assay that follows traditional titer practices where they are serially diluted to provide additional information on the strength of response, the titer value.
Now, nearly thirty years later, the industry still reports ADA values using this approach with titer as the value. However, many people have begun to question if this approach is providing us with the appropriate data to characterize clinically impactful immunogenicity or if it is simply a case of following tradition.
What are Signal-to-Noise (S/N) and Titer for immunogenicity assays?
As described, the traditional 3-tiered approach to ADA testing consists of screening a sample for ADA (screening tier), confirming that positive responses from the screening tier are specific to the drug (confirmatory tier), and titering any sample that screens and confirms positive (titer tier).
Titer is therefore a measurement of the magnitude of the immune response. It is determined through serial dilution, identifying the highest dilution that produces a detectable positive signal in the assay.
Signal-to-Noise (S/N) is a calculation that can be performed using the data from the screening tier. It is calculated as:
S/N = [Equation]
This measurement can be used in place of the Titer value to determine the magnitude of an ADA response in a patient sample.
What are the advantages of S/N over Titer?
- Generates continuous data: Titer assays only provide data as discrete intervals, a positive or negative signal at each dilution, and it is typically only applied to samples that are advanced to this tier after they screen and confirm positive. With signal to noise, every sample tested in the screening tier generates an exact value which could lead to higher precision and better sensitivity for positive responses in the lower range of the assay. This type of data is also better suited for making comparisons between samples from the same patient over time and across analyses.
- Streamlines sample testing: Titer assays are relatively low throughput and consume large amounts of reagents to generate the result. By using S/N, the screening tier can generate the necessary data, eliminating the need for the additional testing tier and saving both time and reagents.
- Provides quality data for clinical interpretation: There have been numerous publications within the industry that demonstrate a strong correlation between S/N and titer values when used for clinical interpretations. Taken together with the other advantages of S/N, this means that a faster, more cost-effective way to measure the magnitude of patient ADA responses can be leveraged to generate data as good or even better than that of the traditional titer approach.
What are the disadvantages of S/N?
- Not every assay is appropriate for S/N: In order for S/N to provide robust, accurate measurements, the assay should have a relatively steady baseline signal from the negative controls, highly specific reagents to prevent nonspecific binding, and should be optimized to prevent hook effect or signal saturation at the upper end of the response curve when possible. If there is a hook or saturation, this phenomenon should be fully characterized to understand the impact on data interpretation as part of decision making. There have been several publications, including a recent one by McCush et al., that provide excellent examples of signal saturation and ways to derive meaningful S/N data despite this potentially complicating factor. However, it is clear that ADA assay format and performance must be thoroughly evaluated to understand if and how S/N can be appropriately leveraged.
- S/N is a ratio, not an absolute concentration: For studies that use ADA data to determine overall magnitude of antibody responses or to characterize trends, S/N is an ideal measurement. Proponents of Titer in the industry have argued that only Titer can provide absolute endpoints, and that this data is required in some cases, negating the use of S/N. However, a standard Titer tier for ADA assays is a semi-quantitative method and does not provide an absolute value either. The base of this argument is likely derived from vaccine titer assays which utilize a global reference standard (such as the WHO International Standards) to calibrate the assay and report endpoint titers as a standardized unit rather than a dilution.
- Validation samples may miss important confounding factors for S/N evaluations: Variability in matrices can significantly impact the background, and thus the negative control values, in an ADA assay. This can be driven by any number of factors such as high levels of drug in circulation or unanticipated interferents in the sample. If these are not accounted for during development, the S/N data generated during patient testing may be at risk. However, this is an issue that would impact Titer values as well. A counter argument for matrix impact on S/N is that the serial dilution of samples as part of a traditional Titer tier results in dilution of potential interferents in the samples which can skew the background signal for each sample in a Titer series and impact the results. By using a single dilution for the screening tier to calculate S/N, you remove that potential bias.
Is S/N accepted as a replacement for Titer by regulatory agencies?
As with all paradigm shifts in bioanalysis, moving from Titer to S/N is a slow process. Within working groups and at scientific meetings, the arguments for using S/N have become the overwhelming majority. In the case of regulatory acceptance, however, this continues to be a case-by-case basis and is a not universally approved strategy or an approach written into guidance. In many successful cases, drug developers have provided rigorous validation data with particular focus on linearity, precision, and robustness of the data compared with titer methods. While S/N is increasingly accepted by the FDA and EMA, sponsors are encouraged to have early discussions with the applicable agencies to communicate the scientific justification behind this approach and ensure an appropriate strategy is implemented.
At Celerion, we work together with our clients to provide the most relevant, scientifically driven, and globally accepted solutions for immunogenicity testing. We look forward to participating in more of these impactful discussions at upcoming industry events and to helping advance the field of bioanalysis in the years to come.
References and Additional Resources:
Goodman, J., Cowan, K. J., Golob, M., Nelson, R., Baltrukonis, D., Bloem, K., et al. (2024). Re-thinking the current paradigm for clinical immunogenicity assessment: an update from the discussion in the European Bioanalysis Forum. Bioanalysis, 16(17–18), 905–913. doi: 10.1080/17576180.2024.2376949. PMID: 39119660.
Lai CH, Chen M, Fraser S, Wang J, McAfee S, Speaks E, et al. (2024) Challenging the Standard Immunogenicity Assessment Approach: 1-Tiered ADA Testing Strategy in Clinical Trials. AAPS J., 27(1):11. doi: 10.1208/s12248-024-00993-9. PMID: 39663329.
McCush F, Wang E, Yunis C, Schwartz P, Baltrukonis D. (2023) Anti-drug Antibody Magnitude and Clinical Relevance Using Signal to Noise (S/N): Bococizumab Case Study. AAPS J. 25(5):85. doi: 10.1208/s12248-023-00846-x. PMID: 37658997.
Stevenson, L. F. (2026). Immunogenicity assays are biomarker assays: is the 3-tiered paradigm fit-for-purpose? An illustrative case study. Bioanalysis, 18(5–6), 503–516. doi: 10.1080/17576180.2026.2677736. PMID: 42199061.
FDA Guidance for developing and validating ADA assays: https://www.fda.gov/media/119788/download