Recommendations and feedback from the European Bioanalysis Forum Workshop: 1 year into ICH M10 – keeping our finger on the pulse

With the new ICH M10 guideline being adopted more than a year ago [1], the European Bioanalysis Forum (EBF) brought together the bioanalytical community in a workshop, “One Year into ICH M10 – Keeping Our Finger on the Pulse” to share and discuss how the guideline has been put into practice. The goal of the workshop was to understand current ambiguities and differences in interpretation by either the industry or the regulators, and from there continue to drive harmonized interpretation and implementation.

The initiative to organize a second workshop dedicated to the ICH M10 guideline in 2023, after a similar workshop was organized last year [2], originates from an EBF members meeting in early spring 2023, where members observed an already growing difference of opinion on the interpretation of various paragraphs of the ICH M10 guideline. As part of EBF's founding goal and mission, they decided to bring together the bioanalytical community to keep a finger on the pulse of the ICH M10 guideline in the early stages of adoption before drifting too far apart.

In line with the EBF's desire to collaborate across all regions, we invited the Japan Bioanalysis Forum (JBF) and the American Association of Pharmaceutical Scientists (AAPS) to participate in or contribute to our workshop. The JBF, which organized a regional ICH M10 meeting in Yokohama in October 2023, shared the outcome of its discussions, which we included as feedback in the respective sessions at our meeting. Unless sporadically referred to, the JBF comments are not discussed in this manuscript but can be found on our website for reference [3]. Within the AAPS, discussion on ICH M10 is ongoing as part of regular Open Scientific Discussion meetings [4]; unfortunately, no consolidated comments from the AAPS were available for presentation during the EBF workshop.

Identifying themes for discussion at the workshop

It is obvious that not every paragraph of the ICH M10 guideline is ambiguous or is being interpreted differently. In essence, the M10 guideline is aligned largely with the regional guidelines/guidance which it is trying to harmonize.

Following up on the decision to organize a workshop, we canvassed our members to identify which paragraphs they felt were ambiguous and required further discussion. For this, we issued a survey which resulted in a long list of about 30 topics for discussion. Some of these topics could easily be addressed, which allowed us to reduce the number of topics to a short list of about 20 (i.e., the ones discussed at the workshop and in this manuscript).

For each of the topics, we then sent a detailed survey to the delegates registered for the meeting to understand and confirm if the topics identified by the EBF indeed were valuable to discuss and what would be the current practice on the guideline requirement on these paragraphs in our community. We received responses from 56 organizations. Hence, the responses can be considered a good cross section of how the community reads, interprets and implements the ICH M10 guideline.

Considering the responses from our delegates, we selected which of the topics required a more in-depth roundtable discussion and which of the topics could be addressed in shorter discussions at the morning plenary or breakout sessions. The agenda of the meeting can be consulted on the EBF website [3].

From early planning onwards and being realistic that it would be impossible to discuss all paragraphs or chapters from the ICH M10 guideline in a 1-day workshop, we decided that chapter 2 (method development), chapter 7 (the paragraphs discussing endogenous counterparts) and chapter 8 (documentation) would not be included in the workshop agenda.

• For chapter 2, we believe the regulatory risk is limited and the industry should stay connected on how we balance anticipated versus real regulatory questions and share those as we move forward. If required, the EBF can include a discussion on method development in relation to ICH M10 at upcoming meetings. Currently, we do not see the industry requesting this.

• For chapter 7, we believe a separate, dedicated discussion may be required, and the EBF may decide to organize a workshop for that purpose.

• For chapter 8, the ambiguities are minimal, but there is a common understanding in the industry that the increased workload to comply with documentation requirements also in earlier development may be disproportional to the value added. The EBF would appreciate the scope of chapter 8 to be limited to late-stage, pivotal studies and/or absolute bioavailability/bioequivalence studies.

In the introduction to the workshop, we reiterated that the intention was not to rehash areas where the industry remains disappointed that the harmonized guideline didn't consider some of the industry's comments during the earlier public consultation phase and for which ultimately the industry hopes that additional data and discussions can lead to some refinements to ICH M10. Also not discussed at the workshop, but as a consequence of the premeeting survey outcome, the industry believes that with ICH M10 the overall cost for method validation and sample analysis has become more expensive (resources, consumables and out of pocket). As industry and health authorities together, it is our responsibility to the patient and the community that this increase in cost adds value.

Dynamics at the workshop

The input from the premeeting survey to the delegates formed the basis on which to build the dynamics of the workshop.

The morning sessions consisted of

• A general session including all delegates, in which the afternoon roundtables related to chapters 1, 5 and 6 were introduced. In this session, the six general roundtables were introduced; then the general items for which we believed a roundtable was not needed were discussed (e.g., incurred sample reanalysis [ISR]).

• Two breakout sessions: 1) with a focus on chromatography assays (i.e., chapter 3) and 2) with a focus on ligand binding assays (LBAs; chapter 4). The breakout sessions had the same structure as the general session. In the chromatography session, the six roundtables were introduced, plus the items for which we believed a roundtable was not needed. These were carry-over assessments during sample analysis, metabolites, choosing the right regression model, hybrid assays and matrix effects. For LBAs, there were no additional themes identified beyond the five roundtable themes.

In the afternoon, the delegates were assigned to roundtables for a more in-depth discussion on the general themes below and could then choose to attend chromatography- or LBA-focused roundtables. For all roundtables, the roundtable moderators traveled from table to table so all could give input in either general + chromatography or general + LBA.

The themes assigned for the general roundtables were the following:

1. Scope interpretation – primary matrix definition

2. Scope interpretation – rare matrix versus tissues

3. Scope interpretation – defining pivotal studies in scope of ICH M10

4. Updating historical validations – when, how and why (not)?

5. Is it allowed to reanalyze positive predose in BE study?

6. Cross-validation – working in the new paradigm

The themes assigned for the chromatography roundtables were the following:

1. Fitting hybrid assays in ICH M10 – day-to-day practice

2. Whole blood stability

3. Analytes and matrices: focus on urine

4. Dilution quality controls (QCs) during assay validation and sample analysis

5. Stock and working solution stability

6. Surrogate/rare/preclinical matrix for chromatography

The themes assigned for LBA roundtables were the following:

1. Dilutional linearity and parallelism

2. Singlicate versus duplicate analysis

3. Surrogate/rare/preclinical matrix for LBAs

4. Chromatography requirements copied into LBA requirements, including tissues/blood stability

5. Dilution QCs during sample analysis

In this manuscript, we provide the conclusions or recommendations from the discussions at the roundtables and from plenary and breakout sessions. More details on the outcome of the surveys as well as the (anonymized) detailed comments given by the delegates can be downloaded from the EBF website [3].

Feedback & recommendations from the workshop

General themes

As mentioned above, the discussions for the general themes related to chapters 1, 5 and 6.

ICH M10 – chapter 1: background & scope

Related to chapter 1 – background

The EBF team reminded the audience the focus provided by the ICH guideline on the 3Rs (replacement, reduction and refinement) [5]. Although the 3Rs are not mentioned specifically in further chapters or paragraphs of the guideline, it is essential to keep the importance of the 3Rs on our radar. The audience agreed it is the ethical and scientific obligation for our generation not to sacrifice animals for blank matrix or any other purpose if there is a valid scientific alternative. It is our responsibility to use all options provided by the ICH M10 guideline and work with the health authorities to get scientific acceptance of all possible replacements of animal matrix with surrogate matrix.

In support of the 3Rs applied more widely to the ICH M10 guideline, the EBF has initiated an intensive experiment to document how surrogate matrix performs in method validation. The results of these experiments will be shared in a separate publication in 2024.

Related to chapter 1 – scope

We highlighted at the meeting that we welcome the clarity provided in the well-written scope paragraph of the ICH M10 guideline. Careful reading should give our community good guidance on which studies, matrices or analytes require full validation and for which an alternative approach can be used.

Nevertheless, as we learned from the premeeting survey responses, a significant part of our community (i.e., up to one out of four members) is unclear on which studies, matrices or analytes require full validation as per the guideline. An even greater part of our community is not considering the guidance provided, and trends toward including more studies/matrices/analytes for full validation than is called for. This is a very important observation which we continue to bring to everybody's attention, because a different interpretation of which studies are in and out of scope will inevitably lead to the industry putting more, if not all, studies in scope of the ICH M10 guideline. Bringing more studies in scope than intended by the health authorities results in an inadequate use of resources, which would be in conflict with the goal of the ICH [6] or the recent 21st Century Cures Act [7]. To further understand the issues or the reasons for ambiguity in our community, we decided to dedicate three roundtables to discuss the specifics of the scope paragraph: definition of primary matrix versus secondary matrix, interpretation of rare matrix versus tissue and, although not specifically mentioned in the scope section, definition of or how to identify pivotal studies.

The recommendations from the three roundtables are as follows:

1. Continue to create awareness on how to define primary matrix versus secondary matrix. Related to the latter, the EBF suggests considering this as “additional matrix” instead of “secondary matrix”, which would aid in the definition. This could be included in Q&A in the ICH documentation. A simple consequence of not being able to define clearly primary versus additional matrix is the inherent risk to make every matrix a primary matrix. A separate discussion was held on urine. The outcome and recommendation from this discussion can be found later in this manuscript.

2. Where the community may interpret the discussion on rare matrix versus tissue to be one related to “when to fully validate”, it is important to highlight and focus this discussion on “when can we use a surrogate matrix?” Indeed, interpretation of primary matrix versus additional matrix will likely, and by default, lead to identifying tissues as an additional matrix, providing the opportunity for a suitably scientific validation approach (e.g., as per earlier EBF recommendation for tissue analysis [8]). The overall recommendation from the discussions at the workshop concludes that it is important to understand the context-of-use in which the data are generated and, if the rare matrix or tissue is not a primary matrix, to apply appropriate fit-for-purpose validation [9] considering the scientific requirements for the matrix and the data required. As suggested by some at the workshop, the EBF may revisit and update its publications related to tissues and fit-for-purpose validations.

3. The label “pivotal study” is only mentioned in chapter 5, ISR. At the roundtable, there was an animated discussion on the word “pivotal” and the impact on our work. Unless for final bioavailability/bioequivalence studies, it is often unclear, if not impossible, for our community to define a pivotal study at the time of sample analysis. An additional point of confusion is that “pivotal” for ISR may mean something different than “pivotal” for a development program, which may even be different from “pivotal” for Health Authority evaluation at filing. Providing more clarity on “pivotal” may be included in Q&A in the ICH documentation.

Updating historical validations – when, how & why (not)?

From the premeeting survey data, we observed two strategies. Approximately half of our industry did not move into a systematic update of historical validations. Consequently, the other half is indeed doing the opposite. Although the EBF cannot judge on the rationale for updating historical validations, we would recommend considering balancing the resources going into unnecessary revalidations and coming together as an industry to define best practices on when to revalidate and for which critical parameters. In this, we assume that in most cases methods were validated toward the validation requirements that were current at the time the studies were performed (e.g., European Medicines Agency [EMA] or US FDA) [10,11] and thes will be adequate for those studies. The EMA is planning a guideline on implementation strategy of ICH Guideline M10 on bioanalytical method validation [12], which was open for public consultation at the time of writing this manuscript.

Indeed, shortly after the EBF workshop, the EMA released the document “Implementation Strategy of ICH Guideline M10 on Bioanalytical Method Validation” to address specific considerations in the implementation of ICH M10 in the European Union. The draft guideline offers a pragmatic stance:

1. If your development program started shortly before 21 January 2023, consider transitioning to ICH M10.

2. If your program is in phase III, ongoing studies may be completed without update, providing the methods were validated against the EMA guideline on bioanalytical method validation from 2011.

3. If you completed the clinical (and nonclinical) development before 21 January 2023 but are submitting your application after this date, there is no need to change or revalidate the bioanalytical methodology according to ICH M10.

A reflection on the pragmatic EMA position above is that the dates for requiring revalidation are evidently aligned with when EMA bioanalytical method validation (BMV) effective dates and those dates may differ when submitting in another region, Those dates may differ when submitting in another region, where either the ICH M10 is not yet in effect or the last regional guideline prior to ICH M10 was in, e.g., 2014 or 2015 (Japanese Ministry of Health, Labour and Welfare [MHLW] or 2018 [US-FDA]). Global alignment may be required to prevent revalidation requirements from being a regional requirement for many years to come.

Reanalysis of positive predose samples in bioequivalence studies

In the first instance, we did not plan for this topic to be a roundtable, although we observed from the preworkshop survey that this question was generating a lot of discussion. Hence, we decided to promote this question to a roundtable, in favor of ISR. A summary of the comments can be found in the slides that were presented at the workshop [3]. Also, during the roundtables, opinions went in multiple directions in favor of or against reanalysis. In summary, and as a recommendation from the discussions, the analysis of positive predose samples from bioequivalence studies should be cognizant of the following considerations: predose samples are not part of pharmacokinetic (PK) evaluation and carry the risk of accidental unblinding. It was also mentioned that if reanalysis is considered, only the positive predose sample from the first period can be reanalyzed.

Cross-validation – working in the new paradigm

Cross-validation requirements as stipulated in the ICH M10 guideline are one of the major differences compared with previous regional guidelines and guidance. Also, during public consultation on the draft ICH M10 guideline in 2019, this chapter generated many questions, as it moved away from generally accepted practices in the bioanalytical community.

As part of the premeeting survey, the continued ambiguity on what is required for cross-validation was confirmed. Approximately half of the responders have not yet updated their processes to the new requirements for cross-validation, either because they are uncertain of how to implement the new requirements or because they have not yet been confronted with the need for a cross-validation. For those who have not yet updated their procedures, they are still evaluating cross-validations against acceptance criteria – that is, mostly the ISR criteria. However, also the organizations that have included the new paradigm for cross-validation are still including similar criteria as an internal beacon to understand the results of a cross-validation experiment. One of the hurdles for a bioanalytical organization vis-à-vis the new requirements is the lack of stakeholder engagement on how to communicate or collaborate when finalizing cross-validations. From the survey, we could read a request for additional guidance with respect to the interpretation of the results (i.e., when should it be done and who needs to do it?). At the 16th EBF Open Symposium, which followed the ICH M10 workshop, a few presentations sharing a process or a case study indicated that there is an inherent risk of more complex statistical procedures creeping into the bioanalytical toolbox; that is, the suggested Bland–Altman plots or Deming regression may not necessarily provide all the answers to interpret a cross-validation [13]. It is the EBF's recommendation to continue the discussion and share examples with the industry and with regulators to ensure we all arrive on, and stay on, the same page.

General themes not discussed in roundtables but limited to the plenary session

In addition to the six roundtables, three other themes were discussed during the general session. The corresponding recommendation or outcome is given in the paragraphs below.

Are method validations in the scope of Good Laboratory Practice?

As part of the discussions in the plenary morning session general items, other recommendations were made. The first one relates to the different interpretation of our industry whether a method validation should be subject to Good Laboratory Practice. From the guideline: for studies that are subject to GLP or Good Clinical Practice (GCP) the bioanalysis of study samples should also conform to their requirements.

Hence, it is our interpretation and therefore recommendation that method validation is not in the scope of GLP. The reason for our recommendation follows previous discussions and agreements within the industry and regulators as documented in the 2014 Organization for Economic Cooperation and Development (OECD) Q&A [14]. ICH M10, being a harmonization document of several regional guidelines on bioanalytical method validation, should be interpreted by our community for GLP as we interpreted the EMA guideline on BMV [11]. Moreover, as per OECD 1, in the scope of GLP are preclinical safety studies [15], and a bioanalytical method validation is not a preclinical safety study. Bringing method validation into the scope of the GLP can be considered to be scope creep, which we should avoid for reasons inherent to amongst others and we repeat the mission from the ICH: to achieve greater harmonization worldwide to ensure that safe, effective and high-quality medicines are developed and registered in the most resource-efficient manner. It is interesting to highlight that the conclusions or interpretation from the JBF is opposite: from its workshop held in Yokohama in October 2023, the JBF recommends that method validations fall under GLP [16].

Biomarkers & immunogenicity

A second recommendation from the morning session relates to biomarkers and immunogenicity. Whereas most of our industry understands that the ICH M10 guideline is not for biomarkers or immunogenicity, a significant group in our industry does use ICH M10 as reference for biomarkers. The EBF is a strong proponent of using the principles of context of use for assays supporting both biomarkers and immunogenicity.

Incurred sample reanalysis

During workshop preparation, ISR was initially identified as a topic for a roundtable. However, when reading the comments, we decided to discuss ISR only at the plenary session. The survey data confirmed that most of the industry is including ISR in more studies than called for by the guideline. This was already the case for the EMA and FDA guideline and guidance, and although the regulators confirm ISR is not required in all studies, the industry continues this practice. Many of the reasons mentioned relate to ISR being seen as an additional quality control or being requested by sponsors. During the discussions in the roundtable on the definition of “pivotal studies,” an interesting insight was shared, in that the way the ISR paragraph is written, the word “pivotal” can be interpreted ambiguously by combining “first” and “pivotal” in one sentence. In this way, there may be more than one “pivotal first study in ... ,” hence calling for more than just the first study. It can therefore be valuable to reach out to the regulators and request clarification on how the word “pivotal” should be read in the ISR chapter. In any case, the EBF stands by its earlier recommendation papers on ISR [17,18] and would continue to stimulate the industry to balance cost with added value of the often-inflated ISR investigations.

Some comments were made challenging the value of the high number of samples required to be reanalyzed for ISR in large bioavailability/bioequivalence studies.

Chromatography themes

For chapter 3 from the ICH M10 guideline, we decided on six themes for roundtable discussion. As for the general themes, also in the themes identified for discussion related to chromatographic assays, some qualified for a roundtable and others were limited to the morning breakout session.

Fitting hybrid assays in ICH M10 – day-to-day practice

The ICH M10 guideline does not specify how to move forward with hybrid assays – that is, immuno-capture LC–MS assays. Within the industry, there is a continued discussion on whether hybrid assays should be considered LC–MS assays or LBAs. At the workshop, about a third of the delegates had experience with and/or knowledge of the principles of hybrid assays.

The audience remains undecided whether more clarity should be provided in the ICH M10 guideline. On the one hand, more clarity in the guideline would be appreciated to remove the fear of not meeting regulatory expectations. On the other hand, overly strict guidance and/or fixed criteria would hamper the scientific freedom which is required, given the complexity of many hybrid assays and the uncertainty regarding the interpretation of certain parameters such as capture antibody saturation or recovery. Discussions in the EBF, including recommendations, in addition to the already published discussion and recommendation papers [19,20], were requested, too. The roundtables gave room for scientific discussions. For more details, refer to the slides presented at the workshop [3].

Whole blood stability

The discussions at the meeting confirmed the premeeting survey data on blood stability testing. The industry is not fully aligned with what is being requested in ICH M10. There was vivid discussion on both the “how” and the “when” at the meeting. As for the previous theme, the roundtables welcomed a scientific discussion which referred to the EBF paper from 2011 [21] as being a good starting point for revisiting the scientific needs. This recommendation paper continues to connect with the scientific details and needs for blood stability testing, and the EBF may want to revise this recommendation in support of providing experimental clarity for blood stability testing in support of the ICH M10 guideline.

Analytes & matrices: focus on urine

The question whether urine should be in the scope of full validation was selected for a more focused discussion in a round table. From the premeeting survey data, we could conclude that it remains unclear for our community if urine is a primary matrix or should be categorized as an additional matrix in studies. At the same time, and even if urine is identified as additional matrix, many still consider urine in scope for full validation. The outcome of the discussions at the roundtables provided clarity in the way that urine is only rarely a primary matrix. Consequently, the EBF recommends that when validating urine, copying the procedures for full validation may not be covering all the scientific requirements. Even though the differences between full validation as per the ICH M10 guideline and additional scientific considerations are minimal, it is important that the community embrace the specific scientific challenges for urine samples (e.g., sampling and storage conditions for high logP compounds or stability of conjugated metabolites, just to name a few). More details are given in the summary slides from the workshop, as well as in an earlier EBF recommendation paper on additional scientific considerations for scientific validation applied to urine samples [9].

Dilution QCs during assay validation & sample analysis

The majority of the audience include dilution QCs at each applied dilution factor with the purpose of process control. One outcome of the discussion was the recommendation to review the scientific requirement of additional validation experiments during sample analysis for lower dilution factors when the higher dilution factor was already fully validated. Because these validation experiments are currently required according to the ICH M10 guideline, the audience agreed to optimize the dilution scheme as much as possible during validation and to apply a limited number of dilution factors during analysis.

Stock & working solution stability

All participants demonstrate the stability of the analyte in solution to cover the duration of use of the solution. However, for working solutions, many do not assess stability when immediately used upon preparation and the remainder is discarded. Since no acceptance criteria are mentioned in the guidance, criteria differ from 5% to 10% or even 15% in the case of large molecules analyzed by LC–MS. The Q&A section of the guideline [22] mentions that two stock solutions can be used interchangeably, provided their content is within 5%. Therefore, many use the same criterion for demonstrating solution stability.

There was a lively debate on the need to demonstrate solution stability for stable isotope labeled internal standards. These materials are expensive and scarce and should be used wisely. Since all include a zero sample in every analytical run to verify the absence of unlabeled analyte and since the same amount is added to every sample within a run and the internal standard is not used for absolute quantification, the general sentiment is that demonstrating solution stability is of no added scientific value.

Surrogate/rare/preclinical matrix for chromatography

As already referred to earlier in this manuscript, the discussion on surrogate/rare matrices continued building on the recommendation from the EBF to maximize the principles of the 3Rs in our industry. The EBF is generating experimental data to provide a scientific stepstone for reducing the use of preclinical matrices wherever possible. Whereas some in our community have embraced such initiatives, others continue to remain uncertain of its regulatory acceptance. We invite other organizations to do the same and have a data-driven discussion with the health authorities on maximizing the principles of the 3Rs.

Chromatography themes not discussed in roundtables but limited to the morning breakout session

In addition to the six roundtables, five other themes were discussed during the general session. The corresponding recommendation or outcome is given in the paragraphs below.

Choosing the right regression model

A valid approach to selection is to define the simple regression model and weighting to be used as standard in an appropriate quality document such as a policy or standard operating procedure. It is recommended that a linear model with 1/x² weighting is utilized as the standard for LC–MS assays. If the validation data meet the predefined acceptance criteria for the validation using the documented standard regression model and weighting, no further action would be required. If the validation data fail the predefined acceptance criteria when using the standard model, it is recommended that the reason for selecting an alternative model and/or weighting be documented.

Matrix effect – special population, hemolyzed & lipemic

Some groups, such as hepatic or renally impaired patients, are recognized as special populations. The classification of other groups, such as different age or ethnic groups or those in healthy versus diseased states, remains unclear. Further discussions revolving around concept of use are proposed.

It was agreed that matrix effects should be validated in hemolyzed and lipemic matrices, as well as regular, to support clinical studies. Ideally, the matrices used should be naturally occurring and representative of the samples. If that is not possible, the M10 guideline provides very clear guidance on how to prepare matrices for the assessments.

Carry-over assessment during sample analysis

Although the premeeting survey showed there was a good understanding of how to manage carry-over during sample analysis, there were different interpretations on how to execute this. At the workshop, several best practice recommendations were discussed on how to prevent carry-over during sample analysis. A summary can be found in the slides from the workshop on our website.

Focus on metabolites

Which metabolites to include in method validation continues to be an area of discussion. Unfortunately, the ICH guideline is not very clear on which metabolites to include for full validation. Consequently, the industry is trending toward including metabolites relatively early in drug development, even as early as the first GLP studies, at the time when it is not clear what would be the added value for systematic quantification of metabolites. In 2016, the EBF published a recommendation paper on when and how to include metabolites in method validation. The recommendation considers guidelines specific to metabolites in safety testing, drug-drug interactions and/or how metabolites contribute to activity [23]. Although from a technical perspective it is relatively easy to include metabolites in validations, the additional workload and costs should be considered and balanced against the added value for the project and the decisions made.

LBA themes

The LBA themes are related to chapter 4 from the ICH M10 guideline. Five roundtables were identified for roundtable discussion.

Dilutional linearity & parallelism

Delegates were aligned in their approach to the assessment of dilutional linearity, with the majority preparing a single ultra-high sample at or above the expected highest concentration and evaluating in a single run with independent dilution series. The results are then used to define the minimum and maximum dilutions that could be applied in sample analysis. The scientific value of the dilutional linearity assessment was questioned by several delegates, as they did not see how dilution of a high-concentration sample across the calibration range differed from the calibration curve itself. For preclinical assays, most delegates prepare samples by first performing the minimum required dilution, followed by further sample dilutions in assay buffer containing matrix (e.g., 10% matrix buffer for a 1 in 10 minimum required dilution), driven by a desire to support the 3Rs and reduce preclinical matrix usage. For clinical assay, where matrix availability is less limiting, there was a greater mix of those who perform dilutions in the same manner and those who dilute in 100% matrix prior to minimum required dilution.

The majority of delegates assess parallelism on a case-by-case basis, typically as a laboratory investigation tool during sample analysis. Several companies assess parallelism the first time that an assay is used in a new matrix (e.g., the first time in healthy subjects or disease-state/indication). This is a similar approach to the (intended) application of ISR, and concerns were raised that this might slide down a similar slope to ISR and be overapplied in areas where it was not merited. The number and nature of samples to be assessed for parallelism were mainly driven by the laboratory investigation. It was agreed that (where possible) at least six to ten samples from different individuals should be assessed, including different concentrations observed in-study (in range to highest concentration). It was discussed that only highest concentration samples may miss parallelism issues, as dilutions to bring the sample into range may dilute out interferences. There were very few cases of nonparallelism, reinforcing the ICH M10 statement that “lack of parallelism is a rare occurrence for bioanalytical methods for PK evaluation.” For the small number of cases where nonparallelism had been observed, the delegates stated that reporting of results could not be defined a priori but rather were defined as a conclusion to the laboratory investigation.

Singlicate versus duplicate analysis

The EBF has put the singlicate/duplicate discussion on its agenda for more than a decade, resulting in a data-driven recommendation paper advocating for the industry to adopt a new mindset and embrace singlicate analysis for LBAs in the regulated environment [24]. Although the recommendation paper already dates from 2019 and the ICH M10 guideline supports singlicate analysis of PK/toxicokinetic samples – as per paragraph 4.2 (i.e., microtiter plates are used for LBAs and study samples can be analyzed using an assay format of one or more wells per sample) and paragraph 4.3 (i.e., the study samples, QCs and calibration standards should be processed in accordance with the validated analytical method) – the majority of the industry hesitates to apply singlicate analysis. The reasons are varied, but common themes are (from contract research organizations) the sponsor does not want to change, internal process, habit, consistency with legacy data or ongoing studies or simply no obvious benefit. From the discussions at the roundtable, we see there is a growing appetite to move into singlicate analysis for new programs in the upcoming years. However, an apparent inborn resistance-to-change gene of the bioanalytical community may stifle a diligent adoption of the new thinking. A few suggestions were made on how the EBF or other communities could assist in helping our community to cross the border on additional sharing of data to showcase that duplicate analysis is something from the past. The good news here is that no additional experimental work is needed, but mining of historic data can provide this information. Singlicate analysis should not be limited to PK/toxicokinetic assays. Although immunogenicity assays are not in scope for ICH M10, it may be good to extend the singlicate/duplicate analysis for those assays, too.

Surrogate/rare/preclinical matrix for LBAs

Although there was much discussion about the possibility of replacing the original matrix, which is considered a rare matrix, with a surrogate matrix, there was no uniform definition of what constitutes a rare matrix. The availability of the matrix was seen as a definition, while cost and adherence to the 3Rs were not suitable as a definition. Consensus was found to use the surrogate matrix early and possibly even start assay development with the surrogate matrix. However, it was agreed to perform comparative runs during development and validation, covering QC recovery, selectivity and dilution linearity in both matrices. Initial results of cross-company experimental data collection based on total ligand binding assay (platform-independent, with or without disruptive assay conditions) indicate the possibility of a successful matrix exchange.

As already referred to earlier in this manuscript and in the Chromatography section, also for LBAs, the EBF is generating experimental data to provide a scientific stepstone for removal of preclinical matrices wherever possible. We invite other cross-company organizations to do the same and have a data-driven discussion with the health authorities on maximizing the principles of the 3Rs.

Chromatography requirements copied into LBA requirements, including tissues/blood stability

Overall, section 4 of the guideline is well written with only minimal occurrences of chromatography requirements coming into scope for LBA where the industry questions the value. However, examples of chromatography requirements within section 4 include extending the calibration range (despite the fact that LBA ranges usually are fixed based on the binding capacity of the capture and detection reagents) and cross-referencing back to section 3 for selectivity and stability.

For the selectivity of LBA methods, there remains ambiguity in the community regarding hemolyzed and lipemic samples and whether they are included as part of the ten required samples or are considered additional samples. While whole blood stability is not specifically mentioned for LBA, such cross-referencing to chromatography brings ambiguity as to whether analyte stability is required in whole blood for LBA methods. Similar ambiguity exists for fixed dose stability requirements in LBA assays. It is the recommendation of the EBF that context of use principles be applied for whole blood stability and fixed dose stability, and appropriate assessments performed only when scientifically valid. The topic of LBA methods for tissues was also discussed within the roundtable. Tissues are rarely a primary matrix; when such methods are not a primary matrix, the application of context of use principles should be the default, rather than applying ICH M10.

Dilution QCs during sample analysis

The premeeting survey and the roundtable discussions assessed the interpretation of the guideline around the inclusion/exclusion of dilution QCs within sample analysis. Although not directly related to the question of the survey, the large majority of the workshop participants responded that the guidance is quite clear in this regard, and they do perform dilution QCs analysis only during stability investigation. However, a minority of the responders (about 20%) are routinely using dilution QCs as process controls during sample analysis. When dilution QCs are used within sample analysis, half of the responders include the dilution factors which were already part of the dilution linearity assessment documented using the method validation. The majority of the responders do not reject the complete run in cases where the dilution QC fails, but only reject the samples with additional dilutions.

In conclusion, there is good agreement among the community regarding not using dilution QCs during sample analysis, as the dilution linearity was already demonstrated during validation.

Additional considerations

As referred to in the introduction of this manuscript, the industry is pleased with a harmonized guideline in which many of the ambiguities were discussed and resolved. At the same time and it is important we keep this on our radar, the ICH M10 guideline unfortunately did not consider some of the comments made during public consultation for which the industry felt that modern guidelines should consider experimental data. It was mentioned at the beginning of the meeting that the workshop was not intended to rehash these items. The meeting focused entirely on ambiguities and trying to resolve these ambiguities during the current implementation phase. At the same time and for those having a longer history in regulated bioanalysis, we should be cautious that we are not in the situation where we were before the EMA guideline on BMV was added to the regulatory toolbox of bioanalytical scientists in 2009 [25]. Indeed, before that time, the industry was referring to a single BMV guideline, the FDA-2001 Guidance on BMV [26]. Already then, the industry was worried that the nonharmonized interpretation by the industry and the regulators of this one guidance was an area of concern. After more than a decade of adding regional regulations [10,11,27–30], and now being full circle with the harmonized ICH M10 guideline being the replacement of these individual regional guidelines, we may be in the situation where different interpretations of this one guideline by the industry and by the regulators bring us back into the same situation as we were in 2008. It is therefore important that we stay together, communicate and share experience and data on how we interpret and implement the guideline.

And it is important we reflect on the following statement: we often refer to the regulatory authorities for raising the bar and requesting unnecessary things, and in the harmonized guideline there are certainly some items where the additional workloads and cost may not be adding value. But the way the industry interprets and implements some chapters/paragraphs of the guideline may also be an illustration of the opposite. In essence, the guideline provides the industry with a clear template on the regulatory requirement for the validation of bioanalytical methods and study sample analysis that are expected to support regulatory decision. The industry should come together not to overinterpret the requirements but to use the tools provided by the regulatory authorities to guide their work in a scientific and resource-effective way. If we fail on this, it may be the industry raising the bar rather than the regulators doing this.

Conclusion

The ICH M10 workshop organized by the EBF preceding its 16th Open Symposium [31], proved to be very useful in the fact that we were able to discuss, share and resolve some of the ambiguities and uncertainties around how we interpret and implement the ICH M10 guideline. We realized from the onset that discussing so many items in a very dynamic breakout setup in only 1 day would be a very difficult endeavor, from the perspective of both the content and the meeting logistics. In this, some of the questions or worries that people brought to the meeting may not have been fully resolved. From here, the EBF is planning to continue the work as stipulated or mentioned in this manuscript by sharing the comments and recommendations and potentially updating some of the earlier recommendation papers to better reflect the new regulatory environment. We also invite other regions to stay connected, as we do observe that different interpretations in different regions may distract from the main goal of the ICH M10 guideline: global harmonized guideline on bioanalytical method validation and study sample analysis. Hence, implementation of the guideline is not an end point but a starting point toward successful harmonized guidelines.