Recommendations and discussion points on immunogenicity, biomarkers, automation/technology and protein–MS from the 2021 European Bioanalysis Forum Focus Workshops

Background

During the first half of 2021, the SARS-CoV-2 pandemic continued to dominate the world, impacting virtually all social interactions globally. Although many of the professionals in our regulated bioanalysis community have received some immunity against the virus either via vaccination or postinfection, we were not immune from travel restrictions. To maintain a maximum of scientific interaction, the European Bioanalysis Forum (EBF) revisited their usual spring meeting calendar and intensified the programming by planning four short Focus Workshops in cyberspace, branded as EBF Cyberconnect events. The themes identified for the workshops were selected from internal EBF discussions and included areas that we felt were of strategic importance and amenable for an online event. As such, and although we continued these discussions within the EBF community, we temporarily paused the outward focused discussions on those themes where we feel the face-to-face connection is (even more) important, for example, ‘Good Clinical Practices for Bioanalysis,’ ‘Risk Based Approach applied for Bioanalysis’ or a continued discussion on the strategic and project based sponsor–vendor interactions. In this manuscript, we will give a short update on the recommendations and discussion points from the four Workshops we organized from March to June 2021: Immunogenicity, Context of Use (CoU) for Biomarkers, Automation/Technology and Protein–MS.

Logistics of the meetings

Although the world has learned to use the technologies for their benefit as imaginatively as possible, having dynamic discussions on science or processes in web-based meetings is not easy. In preparation for their 2020 online meetings, that is, the Autumn Focus Workshop, the 6^th Young Scientist Symposium and the 13th EBF Open Symposium [1–3], we invested significant time to design a tool simulating as much as possible a face-to-face environment, where the threshold to interact was low. Starting points were selecting a stable meeting environment accessible for all at no IT-cost, requesting all speakers to present live from their home or office instead of prerecorded presentations being used, giving ample time for questions and answers during the sessions using an easy chat function and, wherever possible, splitting the audience in ‘video-on’ breakout rooms for discussions with a smaller number of participants. Last, but not least, we ensured all delegates could be identified with name and affiliation, not allowing unidentified acronyms, name abbreviations or telephone numbers to sign into the meeting, so everybody could see directly who was online and also could discuss in private chat if desired. Zoom [4] embedded in a propriety built EBF webpage proved to be a stable and high-performing solution, even acknowledging that some delegates had some issues getting approval to use Zoom on their company hardware. The learnings from the three aforementioned meetings were successfully included in the execution of the four Focus Workshops.

Training day on managing the practical aspects of immunogenicity

On 23 and 24 March 2021, the first out of four EBF Cyberconnect Events was a Training Day on managing the Practical Aspects of Immunogenicity. With experience increasing in industry and with the health authorities, and considering the rapidly growing regulatory requirements, the EBF wanted to share learnings and increase the bioanalytical toolkit to help manage the daily challenges related to antidrug antibodies and provide value-added scientific solutions for immunogenicity questions coming to the bioanalytical laboratory.

The agenda of the training day, which was organized by experts from the EBF Immunogenicity community, aimed to cover as much as possible the different scientific and regulatory challenges encountered when managing immunogenicity questions during drug development, in other words, preclinical immunogenicity, drug tolerance and target interference, outlier exclusion, pre-existing antibodies, neutralizing antibodies, characterization of antidrug antibodies and insights into recent and upcoming recommendations from the EBF on some of the aforementioned areas [5–7]. Special attention was given to understand and correctly interpret the regulatory requirements in which the EBF continues to stimulate a scientific approach as a starting point for interpretation of the regional guidelines. Again, our worries were shared on how emerging regional guidelines (e.g., China NMPA immunogenicity technical guideline) [8] create an even more complex global regulatory environment for which an industry-health authority harmonization discussion, potentially under the umbrella of the ICH, may be valuable. As part of the training goal of the meeting, a full session was dedicated to new modalities and immunogenicity assessment in the context of cell and gene therapy products. The meeting concluded with the review of a number of case studies in which scientific as well as regulatory challenges were discussed.

A clear message from many of the presenters at the workshop was the observation that the industry is at risk of going into a modus of overinterpretation of guidelines or scope-creep, that is, broadening the scope of a guideline beyond its intended area of application, and that continued interactions with the health authorities are key to prevent an inappropriate use of resources to comply with requirements, which were not the intention of the regulators. Industry has a unique opportunity not to fall into old habits of confusing valuable scientific interest with regulatory requirements. It is clear that, although immunogenicity is a well-understood issue for therapeutic (peptides and) proteins, there is still a lot to be learned. We should embrace the opportunity to innovate and use cutting edge technologies and techniques to answer scientific (and regulatory) questions. At the same time, we should be equally attentive to the urge to force the performance and acceptance criteria of those technologies into a regulatory framework that neither was intended, nor designed, for those technologies. Additionally, there may be pressure from management and/or quality assurance (QA) colleagues to be ‘compliant’ to some regulatory guideline intended for other technologies or studies. Instead, we should be confident that a science-led, CoU-driven approach to assay design and validation will result in value-added data that can be defended during regulatory discussions.

The majority of the presentations can be found as downloadable PDF on the EBF website [9].

Focus workshop: biomarkers/CoU: case studies dissected

On 27 and 28 April 2021, the EBF Cyberconnect Events continued with a workshop of further developing the principles to apply CoU as the starting point for method establishment and analysis for biomarkers. After two consecutive Focus Workshop reconnecting the bioanalytical community with the EBF recommendations on biomarker assay establishment and validation and the principles of CoU [10,11], the EBF felt there was a continued and even increased effort needed to prevent biomarker assay validation applying the criteria of pharmacokinetic (PK)-assay validation, in other words, the Bioanalytical Method Validation (BMV) Guidance and Guidelines [12,13]. Certainly with the US FDA BMV Guidance [12] including a paragraph on Biomarkers, for many is becoming the excuse to stay in their comfort zone of their PK-Standard Operating Procedure (SOP) and to apply BMV PK approaches and acceptance criteria to biomarker bioanalysis in all phases of development, irrespective of the decisions taken from the data.

For the workshop, representatives from the EBF Biomarker team built an agenda to stimulate discussions on bringing CoU into practice. For this, we invited experts to share examples and case studies of real-world problems on where applying inappropriate validation criteria or missing out on a priori stakeholder communication on assay requirements and how data would be used, led to a bioanalytical team spending more time and money on inappropriate assay validation or generating results that were not suitable for their intended use. At the meeting, we emphasized the importance of stakeholder interactions and invited end-users of the bioanalytical data to share their view on the EBF recommendations on biomarker assay validation and CoU [14,15]. Indeed, an important stepping stone in the meeting, and the backbone of the first session, were those two EBF recommendations and the feedback from recent workshops by the EBF [10,11] and the American Association of Pharmaceutical Scientists, either as stand-alone workshops or in collaboration.

Certainly the session in which we included the case studies was an eye opener. Listening to the delegates, there is a lot of appetite to implement the CoU principles, but a good portion of the case studies illustrated this is not necessarily the case when it comes to ‘my project’. Feedback from the presenters included a reluctance from either sponsors or contract partners to step away from using a PK-SOP as the basis for validation and analysis, or a misinterpretation of what CoU really implies. Also the difficulty for the bioanalytical laboratory to have the CoU discussion with the stakeholder and end-user remains a major hurdle. In session 3, we discussed how organizational design can make or break a CoU attitude in a company. This session has inspired the EBF Biomarker team to continue their focus in that area and organize additional workshops around the challenges of how organizations strategize biomarker analysis and what are the minimal requirement for an organization to bring CoU into practice.

The panel discussion of the last sessions on ‘value and risk of a regulatory guideline for biomarker assays,’ did not confirm earlier messages taken from other meetings, which suggested that industry wants ‘a biomarker guidance’.

A learning from the panel discussion was a duality which is building in our community expressing two different fears: fear of implementation of guidance for biomarker assays, particularly fear of introduction of a guidance which would be unsuitable for biomarker assay validation, and in the opposing direction, fear of a lack of guidance from the health authorities.

Those that feared the implementation of a biomarker assay guidance, based themselves on the fact that for biomarker assays we need to continue to think scientifically, to consider the different biologic and analytical aspects that can impact the assay that needs to be in place. A prescriptive guidance could not possibly cover the infinite number of CoUs, and risks replacing scientific thought with checklists or tick boxes. For those fearing the lack of guidance, it came down to scientists wanting reassurance that they are ‘getting it right’ and wanting to approach biomarker assays as a tick box exercise rather than engage scientific thought, not knowing what the health authorities are thinking in absence of a formal document, and often driven by intense pressure from management/QA to be compliant to a published (i.e., the PK BMV) guideline.

The desire for some is to have a biomarker assay guidance that incorporates the principles of CoU, for example, with recommendations to have a clear statement of CoU linked to the assay design and validation, perhaps similar to the current EMA guideline for immunogenicity [16], which contains scientific rationale and clear expectations for immunogenicity assay strategies, without being overly prescriptive on the format, design and performance requirements, and considering also immunogenicity assays themselves have different CoU. But there was a recognition that there was simply no ‘one size fits all’ for biomarker assays. With so many studies, analytes, technologies, types of biomarkers, matrices, species, diseases and so many different end-users for biomarker data, the discussions and considerations are often complex (with or without a guidance) for an infinite number of possible CoUs for biomarker assays.

In conclusion, there was clearly an agreement that further discussion is needed - within our community, with stakeholders and with health authorities and that the moment is now. There is a need for additional publications and presentations, for us to share our experiences with peers, stakeholders and health authorities, both on successfully applying CoU principles through drug development and cautionary tales of not applying these principles, coupled with the downstream consequences of these failures. Through workshops and publications, we need to get the word out to all involved: the whole bioanalytical community, stakeholders, including QA and of course the health authorities. With an open scientific mindset, we need to reach out to discuss, educate or even correct them if they are wrong. As we move into the complex new modalities that we are faced with in drug discovery and development, it is upon us as a community to ensure that the message of CoU is clear and implemented. But also we have learned that without guidance, the industry will default to the path of least resistance or at least the path with which they are most familiar with, for example, BMV and PK assay SOPs, the latter not being an evolution we would be supportive of for biomarkers.

An important final message is to ask our regulatory authority colleagues who may be active in writing a biomarker assay guideline, or who are considering developing one: your industry stakeholders would greatly appreciate involvement and interactive discussion prior to a draft version being released, rather than only be given the opportunity to comment on such a draft guideline. There is fear that a draft guideline written without this a priori discussion has the inherent risk of already being too close to requesting PK-like design and acceptance criteria and may not appreciate CoU as the starting point. The delegates at the meeting agreed this was one important take home message.

As for the Immunogenicity training day, the majority of the presentations of the Biomarker workshop can be found as downloadable PDF on the EBF website [17].

Spotlight on (ligand binding & cell-based assays & automation) technology

On 20 and 21 May 2021, the third EBF Cyberconnect Event of 2021 focused on Ligand Binding and Cell-Based Assays and Automation Technology.

The meeting included four themes, either as plenary presentations or in breakout followed by plenary panel discussions:

• Ligand binding assay technological innovation;

• Functional assays becoming ready for the toolbox of the regulated bioanalysis lab;

• Business strategies driving automation or automation driving business strategies?;

• Overcoming challenges – automation solutions delivering value (incl. pandemic inspired solutions).

For each of the sessions, experts from Pharma and CROs shared their experience and vision on where new technologies can add value in drug discovery and development.

The meeting agenda included submitted abstracts from a call for presenters, complemented with feedback from EBF teams working in different arenas of new technologies, for example, the quantitative polymerase chain reaction (qPCR) team or the team working on data integrity and electronic data, both relevant for the meeting. The agenda was very diverse, with significant focus on cutting-edge scientific applications.

For the first two sessions, we included an interactive panel discussion on where the delegates see regulatory challenges on bringing new technologies into the toolbox of the regulated bioanalysis laboratory. Indeed, instrument and software validation requirements have stifled the uptake of new tools in regulated bioanalysis leading to a situation that investments are postponed or we are not able to use these technologies to their fullest potential. In many organizations, instruments are catching dust because they apparently cannot get validated toward not always relevant criteria. Certainly, with qPCR and flow cytometry becoming commonplace tools in quantitative bioanalysis, it is important to keep an eye on the ball and not get distracted by trying to use criteria not required for the purpose where these technologies are used. It is not difficult to see similarities with CoU discussions in the biomarker arena.

The panel discussions used the premeeting input from a survey among the workshop delegates on how their current thinking and processes embrace new technologies and how they fit a regulatory framework (or not). The outcome of the discussions gave a very diverse picture on how the industry is approaching these challenges: most agree that science is the driver of how data are used, and most do tiny, partial or almost full ‘copy/paste’ from the PK-BMV Guideline, leading to anything from manageable processes to self-imposed unachievable acceptance criteria. The outcome of the panel discussions stimulated the EBF qPCR team to increase the pace on their discussion/recommendation paper on processes and acceptance criteria to consider for qPCR (paper in preparation).

On day 2, the sessions on automation were well appreciated. Three (bigger) Pharma organizations showing their strategic vision on what automation in regulated bioanalysis means for them. Given the scale of the investment of the projects presented, it may not be possible for all laboratories to copy the solutions presented, but the vision shared on how thinking out of the box and opening your mind to automate (parts of) processes can also inspire smaller organizations who may not have millions to spend on end-to-end process automation.

The majority of the presentations of this third workshop can be found as downloadable PDF on the EBF website [18].

Scientific & regulatory challenges related to peptide & protein analysis by LC–MS/MS

Finally, between 17 and 18 June, the last of the four EBF Cyberconnect Events from 2021 focused on scientific and regulatory challenges related to peptide and protein analysis by LC–MS/MS.

At the meeting, and with the input from the EBF community and the call for abstracts, we were able to build a strong and diverse agenda, not only focusing on the dominating ‘4-6-15’ (LC–MS/MS) or ‘4-6-20’ (LBA) acceptance criteria question, but including strong science providing food for thought or robust rationale to answer that question.

We included sessions on:

• Scientific challenges: reagents, free versus total, intact versus digested;

• Scientific developments: new tools;

• How to interpret the data: biological specificity versus analytical specificity, which data are ‘true’?;

• Defining a bioanalytical strategy for peptide/protein: which assay, when and why?;

• Regulatory challenges: experience and industry recommendations.

As part of the meeting introduction, we reminded the audience on earlier discussions, meetings and publications from the EBF on the subject [19–21]. More than 10 years ago, the EBF had already organized a symposium on the scientific, technological and regulatory challenges of ‘large molecule’ analysis by LC–MS [19]. Despite the leaps forward in technology, the scientific challenges have hardly changed. This is partly due to a regulatory framework insufficiently recognizing our challenges or not stimulating an open discussion as to how we can take a fresh look at categorizing our assays, but also due to the industry showcasing successes rather than failures, thus painting a picture that peptide and protein analysis by MS is easy and straightforward. A valuable reference point continues to be a recent American Association of Pharmaceutical Scientists publication [22].

As per an EBF recommendation, we repeated our invitation to the industry and health authorities to discuss the value of moving away from technology-driven acceptance criteria toward decision-based criteria, which includes a vision of having only one set of criteria for PK-assays instead of two, driven by the end use of the data and not the technology used to generate the data [23]. In fact, to continue having technology-based criteria in the BMV seems outdated and is creating an increased ambiguity for assays using a mixture of technologies, so-called hybrid assays. With no clear delineation of when an assays should be considered LBA or MS based, the industry is getting increasingly confused and different organizations are sitting on different sides of the fence for identical assays when setting acceptance criteria.

The presentation that perhaps got most attention was from the stakeholder community working with the bioanalytical data and showing the results of a simulation where decisions taken by data obtained from an assay using LC–MS acceptance criteria (4-6-15) were compared with identical data generated by an assay using LBA acceptance criteria (4-6-20) [24]. To our knowledge and even more to our surprise, to date this kind of simulation had not been performed or at least published. In his presentation ‘The Impact of Assay Acceptance Criteria on Derived Data – Pharmacokinetic Assessment Through Simulation,’ Oriol Peris Serrano (Charles River Laboratories) shared one of the conclusions that, using Power Model, the within subject variability, allowing up to ± 20% deviation between nominal and achieved concentrations, did not have an (major) impact on the assessment of dose proportionality compared with the results allowing up to ±15% deviation from the nominal value or without any within subject variability. These data from the simulation confirmed an overall feeling that using either of the criteria will not compromise the decisions taken by the data, be it (patient) safety or efficacy of the drug. At the meeting, we invited the PK and statistician community to build on the work of Oriol. If more confirmation is obtained that 4-6-15 or 4-6-20 yields the same conclusions and does not compromise patient safety, it is worthwhile to open the discussion if more stringent criteria do compromise drug development timelines or inflate costs. In an industry where we have learned to look at what we do from a different angle as part of the SARS-CoV-2 crisis, it is a valuable idea to take a close look at how we are organized around science and some regulatory requirements which may not add the value we thought they did.

The learnings from Oriol’s presentation were still resonating during the closing panel discussion on connecting regulatory and scientific challenges. In preparation for this panel discussion, and in good EBF tradition, the organizing committee had reached out to the EBF community and the delegates of the meeting in two premeeting surveys to probe the industry on strategies they use when deciding on which technology(ies) to use for peptide or protein quantification, which are the acceptance criteria they would be using and how to manage the results coming from complementary or comparative assays, for example, an orthogonal assay applied to the same set of samples. A detailed overview of both surveys can be found on the EBF website [25,26].

The meeting ended with a desire expressed by the panelists and delegates to continue to work together as industry and health authorities to help build an environment that recognizes the value of what the patient needs from the bioanalytical community: value adding data, delivered without delay and without unnecessary burdens on cost and resources. A message that fits the mission and vision of the EBF.

As for all previous meetings, the majority of the presentations of the this fourth workshop can be found as downloadable PDF on the EBF website [27].

Conclusion & future perspective

SARS-CoV-2 clearly limited the possibility of the EBF to bring scientists together in a face-to-face setting. However, as a community we jumped over the hurdles put in front of us and managed four interactive workshops in which we believe to have provided the opportunity for many scientists, over 500 across the four workshops, to share their work and challenges, and provide food for thought and stimulate the discussions on important themes for our community. At the same time, we created an opportunity for the EBF to continue to share our internal discussions with everybody with the aim of informing but also calibrating our vision with additional views: the basis of peer review in science. Hoping the times will change for the better, we hope not to be needing the cybertool in the future, but being realistic we continue to prepare for a world-with COVID rather than a post-COVID world [28], meaning we may use our cybertool for future meetings.