I’d like to thank everyone who attended the recent webinar I gave entitled “What the New AAPS White Paper Recommends for Immunogenicity Assays Supporting Biosimilars”. It was exciting to share more details on the AAPS’ consensus recommendation to use a one-assay approach to develop and validate ADA assays supporting biosimilar programs – explaining how the consensus was reached and exploring best practices the paper puts forth.
The audience was very engaged and I appreciate all of the thoughtful questions that were posed during the Q&A portion of the session. There were so many in fact that I did not get to speak to them all in the live session, and will need to break this follow-up into two parts! I’ve compiled the first set of questions below with my perspectives to share with you now:
What would be the approach if the single assay is not comparable?
First, put on your thinking hat and consider any observed differences in the context of the totality of data collected. Does anything stand out? Second, re-label the biosimilar and originator used as capture/detection reagents in a side-by-side manner under carefully controlled conditions.
Reconsider assay parameters such as the assay platform, MRD, or other critical reagents. Also, obtain and evaluate a panel of multiple positive control antibodies – generated against both the biosimilar and originator – to further an understanding of any observed differences.
It is also important to engage with your CMC team to discuss the analytical characterization of both the biosimilar and originator relative to any observed differences. Differences observed in the ADA assay may reflect previously un-discovered analytical differences between the biosimilar and the originator.
If unexplained differences remain, a discussion with regulatory authorities may be advisable prior to progressing to the next stage.
When selecting the positive control during early development, what should be the acceptable percentage difference between the inhibition of biosimilar and originator?
Development of an ADA assay to support a biosimilar program should employ a systematic, stepwise approach to build up a body of knowledge that instills confidence in the assay’s ability to similarly detect antibodies against both the biosimilar and the originator. Efforts should focus on reducing the potential for variability to allow for meaningful interpretation of assay data.
That said, there isn’t a singular acceptable percentage difference in inhibition. Rather, the paper outlines tangible recommendations for data interpretation and applicable acceptance criteria for data obtained from the antigenic equivalence exercise and comparison of drug tolerance. Any differences should be further considered relative to the observed imprecision of the assay.
If positive control-1 is inhibiting 90% of both biosimilar and originator, and positive control-2 is inhibiting 80% of biosimilar and 60% of originator, which positive control should be selected for further development?
In this case, it may be useful to obtain a panel of multiple additional positive control antibodies – generated against both the biosimilar and originator – to evaluate performance in the assay and better understand this observation prior to selecting and moving forward with any one positive control antibody.
When performing an antigenic assay, must a tracer of both biosimilar and originator be used? Also, what should be the capture/reagent: biosimilar or originator?
To support the application of a single assay approach, it is recommended to use biosimilar-based reagents for both capture/detection in the assay supported with an evaluation of specific parameters – including antigenic equivalence – using both the unlabeled biosimilar and unlabeled originator product(s) as competing antigens for comparison.
Use of biosimilar-based capture/detection reagents ensures that antibodies against the biosimilar are reliably detected, and limits residual uncertainty that the biosimilar is not equally or less immunogenic than the originator.
Could drug competition curves for antigenic equivalence be compared by PLA software?
It is recommended to evaluate antigenic equivalence early in the development phase of an ADA assay to support a biosimilar program. While parallel line analysis (PLA) software may be capable of executing analysis to compare drug competition curves, it may not be the ideal tool for the job.
PLA requires enough data points to adequately define the concentration-response relationship for parameters such as max/min/slope. There may be limited utility in generating a data set of sufficient size at this early stage of assay development. Utilizing PLA software may also increase the risk of detecting differences that, while statistically significant, are not clinically relevant nor meaningful within the context of a qualitative/quasi-quantitative assay.
For these reasons, it is recommended to consider a visual comparison of the drug competition curves obtained for both the biosimilar and originator to ensure they are visually overlapping or comparable, confirming that concentrations of both drug products inhibit assay signal of the positive control to a similar extent. Additionally, a comparison of assay signal at each concentration can provide context to any difference in the signal obtained in the presence of the biosimilar versus the originator, particularly in relation to the estimated precision of the assay (e.g. demonstrating that the %CV of the mean signal obtained from both the biosimilar and originator is less than 20% at most of the drug concentrations evaluated would be indicative of antigenic equivalence).
Let BioAgilytix Support Your Next Project
Thank you again to those that submitted your question and to everyone who joined us for the live session. I’ll be answering the rest of those I received in next week’s blog (read part 2 now), but in the meantime feel free to reach out to me with any additional inquiries about your biosimilar program and how BioAgilytix can support your development needs.
Download the On-Demand Webinar
To access the full recording of the live webinar and Q&A session, click here.