Publication / Source: Bioanalysis 11(15)
Authors: Arnold M
Mark E Arnold, PhD, is Director of Science for Covance Laboratories. In that role, he develops the bioanalytical strategy for immune-, cell-based, quantitative polymerase chain reaction (qPCR) and LC–MS/MS assays to quantify drugs and metabolites, antidrug antibodies and biomarkers in animal and clinical samples for pharmacokinetic and pharmacodynamic assessments. Mark was previously Executive Director of Bioanalytical Sciences at Bristol–Myers Squibb. He received a BS (biology) from Indiana University of Pennsylvania and PhD (pharmacology) from the University of Pittsburgh. For more than 30 years, Mark has been involved in the evolving field of bioanalysis, including the science and the review and interpretation of regulations and guidance. He co-chaired the AAPS Crystal City V and VI Workshops on the US ‘FDA Draft Revised Guidance on Bioanalytical Method Validation’ and ‘Biomarkers’. He is actively involved in the Land O’Lakes Bioanalytical Conference and American Association of Pharmaceutical Scientists (AAPS, named Fellow in 2014). Mark has over 100 peer-reviewed publications, and numerous invited podium presentations.
This interview was conducted by Sankeetha Nadarajah, Managing Commissioning Editor of Bioanalysis.
Keywords: guidance, ICH M10, regulation, scope
Q Can you provide a brief overview of the scope of the draft ICH M10 guidance?
Interpreting the scope as written in the draft M10 guidance has been one of the most challenging aspects for the bioanalytical community in reviewing the document. It does simplify the primary matrix and primary analyte assays as those requiring full validation and permits some metabolites, tissues and additional matrices to use methods with fit-for-purpose (partial) validations. That simplification, however, is countered by the use of the term ‘pivotal’ for studies that have higher requirements. The definition of ‘pivotal’ has evolved over the years and may have its broadest and most impactful implementation in ICH M10. The 2011 EMA bioanalytical method validation guideline used the term in relation to bioequivalence studies; which all agree require the most extensive full validation. The US FDA guidance (2018) expanded ‘pivotal’ to those studies providing data required in support of regulatory decisions (approval, safety or labeling); which from industry–agency discussions seemed to be a subset of all clinical studies and some of the Good Laboratory Practice (GLP) toxicokinetic (TK) studies.
ICH M10 is similar to the FDA guidance, in that GLP TK studies used in regulatory decision-making are included, but expands included clinical trials to ‘all phases of clinical trials in regulatory submissions’. This clause for clinical trials seems to require full validation for the primary analyte/matrix for all clinical studies since all clinical trials are typically included within a regulatory submission. Industry organizations (e.g., American Association of Pharmaceutical Scientists, China Bioanalysis Forum, European Bioanalysis Forum, Japan Bioanalysis Forum) are seeking clarity on which clinical studies require full validation and laboratory execution similar to bioequivalence studies through their comments to their local health authority that are passed on to the ICH.
Q Why was there a need for the ICH M10 guidance, when there is already BMV guidance available?
Differences in regional documents, both when implemented as guidance or as local regulatory law, have made it challenging to perform work that can be accepted globally. Activities to define a sound and globally accepted scientific basis for regulated bioanalysis have been ongoing for nearly a decade through a variety of organizations like those listed above along with the Global Bioanalysis Consortium and Applied Pharmaceutical Analysis. Through initiatives led by these organizations, as well as through bioanalytical meetings – including the Land O’Lakes Bioanalytical and Workshop on Recent Issues in Bioanalysis conferences – regulators and industry counterparts have come together to openly discuss the science and regulations in an effort to define expectations and best practices. Through these efforts, there are much greater levels of communication and transparency for regulated bioanalysis and, while adopting the core practices, each country and region continues to have expectations that are unique.
Combined, these differences have significantly expanded the work needed to develop single bioanalytical packages that can be used globally as part of new drug and generic drug applications; thereby, contributing to increased costs in bringing a new drug to market. Achieving an ICH guidance in M10 that will be accepted in so many countries and regions will establish a common set of expectations, reduce some of the unique differences and allow for the more efficient generation of data for new drug approvals for patients.
Q Can you explain the differences between a full, partial and cross validation, and how you decide which one is appropriate to specific studies?
The key thing to remember with what I am about to discuss is that this is based on the current draft guidance and is subject to change. As mentioned before, full validations are required for assays of the primary analyte(s) in the primary matrix for TK studies and pivotal clinical studies (including bioavailability [BA]/bioequivalence [BE]) that provide data that support regulatory decisions. That places some metabolites and additional matrices in the category of requiring only a partial validation. The data from these assays would help characterize the presence of the drug but would not be essential to a regulatory decision. One interesting aspect is that partial validations may also be used if a fully validated assay is used for another matrix or the same matrix in another species. Questions have been raised as to whether this permits, for example, a rat plasma assay used to support an investigational new drug (IND) toxicity study can be partially validated for an IND-supporting toxicity study in dogs. Another use for partial validations is for a demonstrating that minor changes to a fully validated method have not introduced impactful changes to the method.
Cross validations are another area of the draft guidance in which there are several options. In this case, however, the requirements are to ensure that the data generated within a study or between studies may reliably be compared. The three cases defined within M10 are: when two methods are used in a single study; when different methods are used in different studies; or when samples from a study are analyzed in two labs using the same method. Industry scientists are in agreement that the first and third situations make sense but several reasons have been discussed for not endorsing the second. One key challenge is that intersubject variability is typically higher than assay variability, rendering comparisons of the methods less valuable. Additionally, the current drug development paradigm often involves transfer of a drug between companies (through direct purchase of the drug or company mergers) or running the assay in multiple labs, both resulting in multiple labs and assays being used over the course of the development program. It is difficult, therefore, to interpret the relationship between the earliest and final studies across the series of cross-validations that would be required in such a scenario.