Challenges of analytical method validation for ATMPs

Quality control of medicinal products is a fundamental aspect of assuring the quality, efficacy, and safety of the product. Testing of the product at intermediate manufacturing stages, such as testing of Drug Substance (DS) and on the final Drug Product (DP), are performed to confirm the product meets the established specifications as defined in the Marketing Authorisation.

Analytical Target Profile

The attributes of the DS or DP that are to be tested to obtain the necessary information about product quality safety and efficacy are determined during the Drug Development stage.

The information gathered from the test results supports the establishment of the Analytical Target Profile (ATP), which provides a required range in the quality criteria for key performance characteristics of the product. Ranges defined in the ATP create a direct line from the product’s Critical Quality Attributes (CQAs) and the release specifications. Establishing the ATP for each assay in relation to the CQAs in the early phases of product development should be a priority for any drug developer, to provide clarity and direction for the ongoing analytical and process development. A timely start at the early development phase also avoids non-productive development, saving time and money.

Establishing these product and analytical characteristics, and measuring them consistently, helps to support regulatory submissions and claims of comparability at later phases. As the product develops and is characterised, the associated analytical methods also move from the Research and Development (R&D) stage into Good Manufacturing Practices (GMP) manufacturing. From the Clinical Trial phases onward, analytical methods are routinely used, monitored and subject to continual improvement that are managed via Change Management. This is also known as the Analytical Product Lifecycle, and guidance in the analytical procedure development and its lifecycle is given in a new draft document from the International Council for Harmonisation (ICH); ICH Q14 “Analytical Procedure Development”.

Analytical method validation

Once a product is manufactured in accordance with GMP, the analytical test methods are required to be validated and described in the product’s Marketing Authorisation or Technical Dossier. Validation of these methods can be a complicated venture, depending on the type of analytical method and product. International Council for Harmonisation (ICH) guideline ICH Q2R1 ‘Validation of Analytical Procedures: Text and Methodology’ provides guidance that Quality Control Laboratories can avail of in consideration of which analytical method characteristics are applicable for method validation.

Despite the information available, validation of analytical methods for Advanced Therapy Medicinal Products (ATMPs) can be even more challenging. This is due to the inherent variability in starting materials, complex biological features, and associated manufacturing process. Other factors are the limited batch history and sample availability, due to low batch sizes and high manufacturing costs, and lack of available assay references and assay controls.

Facing challenges for ATMPs

Unlike more mature biological pharmaceutical products, such as monoclonal antibodies, which like ATMPs such as AAVs are highly heterogeneous, analytical techniques and process manufacturing workflows are still the subject of intense development optimisation and change. The analytical methods used to support ATMP drug development and release can be thought of in terms of three categories related to their level of maturity and expected analytical development and validation time costs. Firstly, ‘fully mature’ drug characteristics that are analogous to those seen in more established molecule types can be analysed using established methodologies employed for other biological molecules. Examples include excipient testing for antifoam clearance, and Host Cell Protein (HCP) and Host Cell (HC) DNA impurity. This makes them relatively simple to implement from the earliest phases of development, often with kit-based assays and fully GMP analytical systems and software.

A second group of assays includes those which are commonly found in GMP QC release laboratories. Therefore, they can take advantage of the fully developed analytical platforms and software used for this purpose, but with some additional considerations that may require further analytical development effort and special attention during validation. Examples in this second group include the measurement of characteristics such as Post Translational Modifications (PTMs) of capsid proteins by peptide mapping, analysis and relative quantification of capsid proteins by LC-UV/FLD or CE UV/FLD, and SEC for aggregate analysis, and protein impurity determinations can be expected to require further development if based on equivalent assays such as those used in mAb products.

PTMs measured at the peptide or protein level must consider the likely impact of sample preparations used to disrupt the capsid. SEC and other particle sizing methods have to be suitable for larger size monomers and aggregates, with AAV molecules being >30 times larger than monoclonal antibodies and protein concentrations also likely to be significantly lower, and thus sensitivity may be a consideration. Monoclonal antibodies are often manufactured to concentrations greater than 100mg/mL, whereas AAVs are likely to have protein concentrations of <0.05mg/mL and are manufactured at significantly lower batch volumes. Protein impurity assessments are also likely to be required to perform at protein concentrations significantly less than those seen for antibodies methods and may be required to perform at a level too low for many existing procedures.

To continue with AAV molecules, several characteristics such as empty/full assessment, infectivity and, potency rely on techniques that are uncommon in pharmaceutical release settings. By their nature, AAV molecules are highly complex and therefore methods to measure this complexity may require significant development and optimisation in order for them to be appropriate for use in a GMP setting. As a result, this type of method analysis could likely change the most during the development process, leading to significant extra work in proving comparability of the analytical method over the course of the assay lifecycle.

It can also lead to uncertainty during specification setting in early phases. Techniques such as AUC and cryoEM would be more than capable of evaluating empty, full, and partial AAV species, but are not routinely used in a GMP setting. Therefore, these methods do not have commercially available Annex 11 or 21 CFR 11 compliant software that could be used as part of the system’s qualification in a GMP environment. This can lead to significant time investments for later validation activities closer to the commercialisation stages of the product.

Product availability and reference standards

As described previously, the amount of material that is manufactured is often in very small quantities. In some cases, these bespoke medicines are limited to just the amount that serves as the dose given to the patient. Due to the novelty and complexity of ATMPs, reference materials are not always available, thus making it more difficult to demonstrate that the procedures and tests developed are suitable for their intended use. The use of interim references can be an option in these circumstances and their use could be recommended to provide a level of continuity and confidence in the analytical methods applied; in particular, as these interim references are developed through their analytical lifecycle where this is a requirement. Reference standards should be representative of the manufacturing process as much as is possible and may need to be replaced with associated bridging studies as the process develops. Where reference standards are not available, in-assay controls can help to demonstrate assay consistency and support in proving representativeness during the drug development lifecycle.

Data from reference standards and analytical controls should be used to set assay acceptance criteria. These criteria may also be wide compared to other modalities at the clinical phase stage. However, as the drug development progresses, the results from these assays should also provide assurance of manufacturing consistency. Trending results over time should confirm release criteria and, where assay variability is present, studies should be performed to ascertain the cause and establish understanding of inter- and intra-assay variability. In line with ICH Q14, Design of Experiment (DoE) studies could be useful in designing these studies and can help conserve limited sample amounts.

Regulatory bodies

Throughout the development process, it is recommended that continued and frequent dialogue with regulatory agencies is maintained. Communications should include details of draft analytical strategies. This should help focus efforts for continued analytical development and validation. It also builds an understanding and justification for areas where sample and data availability may be limited. With agreement from regulatory agencies, it may also allow for the gradual reduction of release criteria once it is clear that these criteria are either not relevant to safety and efficacy, or not required due to process robustness such as with process excipients.

Assay comparability studies

The storage and prudent use of retain samples from all key process lots could likely be critical in establishing assay comparability or for analytical bridging studies through the development process; notably, where assays require significant development and processes have undergone significant change, as is likely with these ATMP product types. In addition to these studies, the development of platform approaches to all analytics should be considered. Although these technologies are emerging, data from similar molecules, for example of other serotypes in gene therapy, can be leveraged to support method development, validation, and specification setting activities, where appropriate.

Despite the fact that release analytics for new modalities may have unique challenges - due to the use of new techniques, a lack of historic data, and an increasingly complex mode of actions - this is not the case for the majority of assays used for release of these potential new products. A significant proportion of the required analytics are either compendial or based on well-established procedures. Time and effort can more easily be focused by following a systematic approach, as defined in ICH Q14. Appropriate utilisation of experience and understanding will allow the generation of a suitable analytical strategic plan to develop methods appropriately during the process development, which will be able to ensure the quality, safety, and efficacy of the product.

About the authors

Patrick Nieuwenhuizen is director senior consultant at PharmaLex. With a microbiology and sterile manufacturing background and over 25 years’ experience in the pharmaceutical industry, Nieuwenhuizen has worked across a variety of platforms, including biologics, sterile fill finish and solid oral dose. In addition to site responsibilities, he is involved in several corporate initiatives such as the Sterility Assurance Council and the roll-out of corporate standard programmes. Additionally, he has acted as a lead auditor, involved with audits facing several competent authority inspections including but not limited to the HPRA, FDA, ANVISA, Chinese FDA, and Canadian Health Authority inspections.

Christopher Rogers, MSc, is senior manager of regulatory affairs, CMC, at PharmaLex UK, managing delivery of regulatory projects and contributing to a range of ATMP and biosimilar regulatory projects. With 20 years’ experience in biologics development and manufacturing, focusing on routine and advanced analytics from characterisation through to analytical development and quality control for CMC, he now supports regulatory CMC. Rogers holds a BSc in Biology from the University of Hull and an MSc in Analytical Bioscience and Drug Design from the University of Salford. He is a keen cyclist.