Robust Control Strategies for AAVs: Use of QbD within the Regulatory Frameworks
The development and manufacturing of gene therapy products, particularly those based on Adeno-Associated Viruses (AAVs), present unique challenges in ensuring consistent product quality, safety and potency.
As the field of Advanced Therapy Medicinal Products (ATMPs) continues to evolve, regulatory expectations are increasingly shaped by established frameworks such as the International Council for Harmonization (ICH) guidelines and ATMP guidelines of the U.S. Food and Drug Administration (FDA), United States Pharmacopeia (USP), European Medicines Agency (EMA) and other relevant regulators.
Particularly the ICH Q8 (Pharmaceutical Development), Q9 (Quality Risk Management), and Q10 (Pharmaceutical Quality System) give good orientation during the development of the manufacturing process and of the cognate control strategy. In Biologics and Vaccines these concepts were established and applied successfully for decades. Due to the complexity of the ATMP's manufacturing process and final products, the comparatively still limited CMC (Chemistry, Manufacturing, and Controls) knowledge base is reflected in the frequency of regulatory objections and application rejections1.
The ICHQ10 defines the control strategy as a set of set of controls derived from product and process understanding assuming process performance and product quality.
Five core critical quality attributes can be assigned: identity, purity, safety, content, and potency, ensuring product quality, efficacy and patient safety.
There are two major differences comparing the manufacturing process of viral products with the biologic-based ones:
- Its manufacturing scale
- Its number of process steps
The high viral yield of these virus manufacturing cells and small patient number for these rare diseases allow to work with lab/small scale at USP & DSP. With a small-scale outlay of the manufacturing process, one can already meet the cognate medical needs. Further, the in-vivo gene therapy (GT) process consists of a smaller scale and of less process steps in comparison to biologics. Despite the complexity of the recombinant Adeno Associated Virus (rAAV) product and its impurity profile there are less process steps needed. Please find an illustration in Figure 1.
Figure 1: Comparison of suspension-based manufacturing process between mAb (A) and AAV (B) from USP to DP
Abbreviations: AAV Adeno associated virus, AEX Anion exchange Chromatography, AFC Affinity chromatography, C Centrifugation, CEX Cation Chromatography, CHO Chinese hamster ovary, CL Clarification, DP Drug Product, HEK Human embryonic kidney, DF depth filtration, mAbs monoclonal Antibodies, SF Sterile filtration, TFF Tangential Flow Filtration, UF Ultrafiltration VI Viral inactivation
Recommendation
Bernried, Germany10/11 March 2026
Low Endotoxin Recovery/Masking - Hands-on Laboratory Training Course
Beyond the ICH framework, both the EMA and the FDA have issued additional guidance specific to gene therapy products. These include general frameworks as well as more targeted documents - such as EMA's guidance and the FDA's draft guidance on potency testing - highlighting the growing regulatory emphasis on demonstrating product functionality and consistency.
The principles of Quality by Design (QbD), as outlined in ICH Q8, Q12 and in particular Q14 play a pivotal role in the development of AAV products. By integrating risk assessments throughout the drug development lifecycle, QbD supports the optimization of both manufacturing processes and analytical methods, ultimately contributing to a more predictable and controlled product profile.
This article explores how these regulatory and scientific principles converge to shape the control strategies for AAV-based gene therapies, with a focus on aligning CQAs by use of QbD with evolving expectations and facilitating successful product development and approval.
The complexity of AAV-based gene therapy products demands highly robust, reliable, and well-characterized analytical methods. Quality by Design (QbD) offers a solution with its systematic, science- and risk-based approach that enhances method understanding and control throughout the product lifecycle.
Based on the Quality Target product profile (QTPP), the resulting Analytical Target Profile (ATP) is a clear statement of the method's intended purpose and performance criteria. For AAVs, this includes quantifying viral genome titer, assessing capsid integrity and measuring potency. The ATP sets the foundation for method design and validation, ensuring alignment with product CQAs.
Critical Method Attributes (CMAs) are the measurable characteristics of an analytical method that must be controlled to meet the ATP. For example, in a qPCR assay for genome titer, CMAs might include amplification efficiency, specificity, and linearity. Understanding these attributes helps prioritize method parameters that influence performance.
Using tools for risk assessment and method design such as Ishikawa diagrams, Failure Mode and Effects Analysis (FMEA), or Design of Experiments (DoE), developers can systematically evaluate the impact of method parameters (e.g., reagent concentration, temperature or pH) on CMAs. This step is crucial for identifying Critical Method Parameters (CMPs) and understanding their interactions.
DoE enables efficient exploration of the method design space, allowing developers to optimize CMPs and define a Method Operable Design Region (MODR)-a multidimensional space within which the method consistently meets the ATP. Robust testing ensures the method performs reliably under small, deliberate variations.
QbD does not end with method validation. Under ICH Q12, analytical methods are subject to lifecycle management, where performance is monitored over time using trending tools and control charts. This proactive approach supports continuous improvement and facilitates post-approval changes with a reduced regulatory burden.
Benefits of QbD in AAV Analytical Development are:
- Enhanced method understanding and control
- Improved regulatory compliance through structured documentation and risk management
- Greater method robustness, reducing variability and out-ofspecification results
- Facilitated technology transfer and scalability across development stages
Recommendation
Bernried, Germany12/13 March 2026
Monocyte Activation Test (MAT) - Hands-on Laboratory Training Course
In the field of gene therapy, quality management based on Quality by Design (QbD) principles (ICH Q8 and ICH Q12) has gained increasing importance, particularly at the molecular level. As the drug product consists of a viral packed nucleic acid template and the active molecules are synthesized within the intended target cells. An additional aspect unique to the gene therapy field is its cellular dimension: the QbD framework extends not only to the nucleic acid "template" itself but also to the cellular processes underlying the biosynthesis of the active transgenic molecules.
Consequently, the correlation between these intracellular biosynthetic processes and the relevant quality attributes within the target cells must be considered simultaneously.
A proposal of a cellular distinction of Quality Attributes of recombinant AAV based on triple transfection/plasmid manufacturing is shown in Figure 2.
Figure 2: Cellular Quality attributes of recombinant Adeno Associated Virus (rAAV)
(A) Extracellular attributes on helper plasmid and (B) Intracellular attributes on molecular level (DNA & mRNA)
Abbreviations: Cap capsid, ds double stranded, ITR Inverted Terminal Repeat, PR/p promoter, Rep replication, ss single stranded, UTR untranslated region
Therefore, in this case the term QbD is even literally understood and requires a holistic approach considering processing steps of the cellular- and bioprocess and its combined impact on clinical efficacy.
Any potential quality gaps on molecular design level may impact cellular processing steps and subsequentially may impact clinical efficacy. 3 key intrinsic features derived from the nucleic DNA template for GT products are:
- Motifs and primary sequences (cellular factors binding motifs, expression regulation, expression, pathogen binding motifs, immunogenicity, packaging, others)
- higher order structures of the template DNA and of its primary products (mRNA/RNAs)
- Functionality of the transgenic payload (expression, functionality, potency)
Over the last decades In Silico analysis became crucial tool of drug discovery for the identification of potential targets2, 3, 4. In proteomics numerous tools and applications were driving the successful development of protein-based drugs and its medical translation. Another promising application are Multi-omics/in-silico models consisting of genomics, transcriptomics, proteomics and metabolomics as data collection tools. They help to improve rAAV manufacturing process by analyzing specific producer cell lines after transient transfection4, 5.
On the ATMP side, for instance AAV based therapeutics, an increasing product understanding, improved manufacturing processes in regard to purity and yield and new analytical approaches and technologies, deepened the knowledge of the cognate cellular quality attributes. These progresses contributed to the sequential modulation of the specific motifs and primary sequences which are involved in cell entry, promoter regulation, transcription, expression, packaging, processing and trafficking, immunity6, 7 ,8 ,9 ,10.
More precise predictions of the intracellular processing depending on the transgenic sequences (expression cassette) would allow better modification of corresponding cellular quality attributes (Fig. 2) which will help to reduce or avoid product related impurities and product variants4, 7, 11.
Another crucial intrinsic factor is the higher-order structure of the DNA template and its primary products, which strongly influence cellular attributes (Fig. 2). Identifying the nucleic acid sequences responsible for such structures, and understanding their potential formation, may enable prediction of these higher-level folds of the gene therapy template and their primary RNA product. Such appropriate computational tools are essential and applied already on 'secondary structure level12, 13 and on tertiary structure level14, 15. Except for a few biological RNAs, the most common understanding of quaternary nucleic acid structures is DNA/RNA-protein complexes16, 17, 18.
In general, a stronger consideration of patient-centric quality aspects like for vaccines is recommended during development and selection of high-reliable prediction tools19, related tests for its statistical significance and robust datasets20. Sometimes also experimental verification might be needed21. The impact of potential modifications/ alterations of the transgenic sequences may be handled by risk assessments like ICH Q9 and according to measures described in ICH Q14.
A harmonized approach among the industry with respect of standardized tools, design and usage of corresponding data banks is desirable. Investment in quality on molecular level prevents later labor- and cost intensive efforts in CMC manufacturing.
In the context of development project timelines, an optimal time point for quality check of the three intrinsic key features might be during the lead selection prior going to Toxicology studies.
The increasing usage of AI and continuous development of prediction tools might make the estimates more reliable22.
Investments in the improved cellular quality on molecular level can lead to significant cost reductions by leaner manufacturing, a leaner control panel, an improved safety profile and a potential higher efficacy.
There is not only the need for more specified regulatory documents and guidelines with respect to damaging sequences/motifs as well as safety & efficacy relevant sequences but also for harmonized concepts and standards which consider the design and handling of these cellular quality attributes.
Briefly said, in contrast to biologics and vaccines the timepoint of the definition of the product quality in the case of GT based products (like AAV) is already during the preclinical phase. Main driver for this shift is the molecular design of the transgenic nucleic acid sequence. In addition, also shorter timelines to Entry to Human (Phase one) impact the CMC activity and related product quality. Therefore, an earlier consideration of these CMC development aspects during preclinical stages will lead to a smooth transition to later clinical phases and GMP.
About the Authors
Dr Ulrike Herbrand joined Charles River Laboratories in 2007. She is Scientific Director Global in vitro Bioassays and Head of the Bioassay Research & Development team at Charles River Laboratories' site in Erkrath, Germany. She is an expert in mechanism of action-reflecting bioassays for protein therapeutics and ATMPs.
Dr Roland Pach has been the global CMC Analytical Technical Lead in the cancer vaccines and cell- & gene therapy (CGT) area of Roche more than 10 years. He represents Roche in external development projects and numerous due diligences of in-licensing candidates or companies in the CGT fields.
Note:
1 Barkholt L et al., European regulatory experience with advanced therapy medicinal products. Nat Rev Drug Discov 18, 8-9 (2019)
2 Murray, D, Doran, P, MacMathuna, P. et al. In silico gene expression analysis - an overview. Mol Cancer 6, 50 (2007)
3 Miah Roney et al., The importance of in-silico studies in drug discovery, Intelligent Pharmacy 2, 578-579 (2024)
4 Fu Q, Polanco A, Lee YS, Yoon S, Critical challenges and advances in recombinant adeno-associated virus (rAAV) biomanufacturing. Biotechnol Bioeng.120, 2601-2621 (2023)
5 Esse R et al., Multi-omics data integration to improve adeno-associated vector (AAV) manufacturing. Cytotherapy 26, S133 (2024)
6 Large EE et al. Adeno-Associated Virus (AAV) Gene Delivery: Dissecting Molecular Interactions upon Cell Entry. Viruses 13, 1336 (2021)
7 Maurer AC, Weitzman MD. Adeno-Associated Virus Genome Interactions Important for Vector Production and Transduction. Hum Gene Ther. 31, 499-511 (2020)
8 Dhungel, BP et al., Journey to the Center of the Cell: Tracing the Path of AAV Transduction. Trends in Molecular Medicine 27, 172 - 184 (2021)
9 Srivastava A, Rationale and strategies for the development of safe and effective optimized AAV vectors for human gene therapy.Molecular Therapy - Nucleic Acids 32, 949-959 (2023)
10 Keeler AM er al., The curious case of AAV immunology. Mol Ther. 33, 1946-1965 (2025)
11 Xie J et al., Short DNA Hairpins Compromise Recombinant Adeno-Associated Virus Genome Homogeneity. Mol Ther. 25, 1363-1374 (2017)
12 Afzal M et al, RDNAnalyzer: A tool for DNA secondary structure prediction and sequence analysis. Bioinformation 8, 687-90 (2012)
13 Sato, K, Akiyama, M & Sakakibara, Y RNA secondary structure prediction using deep learning with thermodynamic integration. Nat Commun 12, 941 (2021)
14 Kagaya, ., Zhang Z, Ibtehaz, N et al. NuFold: end-to-end approach for RNA tertiary structure prediction with flexible nucleobase center representation. Nat Commun 16, 881 (2025)
15 Zhang Y, Xiong Y, Yang C, Xiao Y, 3dRNA/DNA: 3D Structure Prediction from RNA to DNA. Journal of Molecular Biology 436, 168742 (2024)
16 Jones CP, Ferré-D'Amaré AR. RNA quaternary structure and global symmetry. Trends Biochem Sci. 40, 211-20 (2015)
17 Li, J., Chiu, TP. & Rohs, R. Predicting DNA structure using a deep learning method. Nat Commun 15, 1243 (2024)
18 Marklund E, Ke Y, Greenleaf WJ. High-throughput biochemistry in RNA sequence space: predicting structure and function. Nat Rev Genet. 24, 401-414 (2023)
19 Mathur G, Pandey A, Goyal S. A comprehensive tool for rapid and accurate prediction of disease using DNA sequence classifier. J Ambient Intell Humaniz Comput., Jun 25:1-17 (2022)
20 Magnus, M. et al. RNA-Puzzles toolkit: a computational resource of RNA 3D structure benchmark datasets, structure manipulation, and evaluation tools. Nucleic Acids Res. 48, 576-588 (2020)
21 Zhang, J., Fei, Y., Sun, L. et al. Advances and opportunities in RNA structure experimental determination and computational modeling. Nat Methods 19, 1193-1207 (2022)
22 Chaturvedi M, Rashid MA, Paliwal KK. Transformers in RNA structure prediction: A review. Comput Struct Biotechno J 27, 1187-1203 (2025)
