Cell and gene therapies (CGTs) are a novel and fast-growing class of transformative therapies designed to address gaps in traditional treatment strategies of some of the most severe diseases. By definition, gene therapy “seeks to modify or manipulate expression of a gene to alter the biological properties of living cells for therapeutic use” (1). That can be either an in vivo delivery of a gene or delivery of a gene to a patient’s cells that are manipulated outside of the body and then redelivered, which is then considered a cell-based gene therapy.

In recent years, several CGTs have been approved for market. Such products include Kymriah (tisagenlecleucel, Novartis) and Yescarta (axicabtagene ciloleucel, Kite/Gilead), which are autologous T cells engineered to express chimeric antigen receptors (CARs) directed against CD19 for treatment of B-cell lymphoma; Luxturna (voretigene neparvovec-rzyl, Spark Therapeutics), which is an adenoassociated-virus–based gene therapy for the correction of RPE65 in treating blindness; and Zolgensma (onasemnogene abeparvovec-xioi, AveXis/Novartis), which is another adenoassociated-virus–based gene therapy for the correction of SMN1 in the treatment of spinal muscular atrophy. As the market continues to grow, some analyses project that by 2025, 10–20 CGTs will be approved every year (2).

Although CGTs are unique compared with traditional biopharmaceuticals (e.g. therapeutic proteins and monoclonal antibodies), manufacturing strategies for CGTs have piggybacked on learnings from biopharmaceuticals. However, poor understanding of the mechanisms of action (MoA) and the added complexity of defining living cells as therapies, have made properly defining CGT products difficult, especially when developing appropriate release assays that ensure both safety and efficacy. One approach to reducing the risk associated with such therapies is to build a manufacturing process that includes validated analytical procedures and assays to reduce variability and ensure high-quality and consistent products.

Manufacturing CGTs Using a QbD Approach
Implementation of a manufacturing process that assures a predefined quality of product is a critical requirement for the licensing and marketing of every CGT product. Generally, extensive development is needed to mature an early stage process to commercial manufacturing because of the biological complexity of the products. However, prioritization of speed to early clinical trials — where the focus is on maintaining supply of clinical materials and aversion to the risk of early overcommitment for unproven products — can lead to limited opportunities for process development (3). Inadequate process knowledge and understanding constrict implementation of process changes because the impacts on product safety and efficacy are unknown. That often leads to the adoption of processes that, although compliant with established regulations, are not optimal for assuring broad availability to patients who depend on those therapies.

To improve manufacturing of CGTs, a quality by design (QbD) approach can be adopted in which knowledge of a product and risk assessment are used to manage the process and build specifications around products (4). Manufacturing a CGT product should start with defining a quality target product profile (QTPP) that includes all product properties that are desired for the final product and ensures its high quality, safety, and efficacy.
A risk assessment then is conducted to determine critical quality attributes (CQAs), properties that are directly connected to product efficacy and safety and that should be monitored, assayed, or assessed during manufacturing to ensure that the QTPP is met (5). Another risk assessment is performed to identify critical process parameters (CPPs) that affect CQAs. The effects of CPPs are quantified in a design space, typically by using studies enabling design of experiments (DoE) and analytical assessments implementing systems modeling. That knowledge is used to develop a control strategy for maintaining CPPs within operating ranges that ensure consistent performance of a process with respect to established CQAs.

The control strategy needs to be validated, continually monitored, and improved as needed, especially as knowledge and understanding of CQAs increase over time. Before commercial production can take place, process performance qualification (PPQ) manufacturing runs are conducted at commercially relevant scales using qualified facilities, utilities, equipment, and personnel to demonstrate consistency in manufacturing within predefined operating parameters (6).

PPQ runs usually are scrutinized more than manufacturing runs in terms of sampling, testing, and monitoring. That knowledge is used to determine the sampling frequency of commercial batches. For processes in which inline or real-time measurements of CPPs affect process changes, the focus is on qualifying the measurement systems and control loops for measured attributes. The overall goal of PPQs is to qualify designed manufacturing processes as capable of reproducible commercial manufacturing. For process validation, scientific evidence that a process can consistently produce a predefined quality of products is established by gathering and analyzing data from process design through commercial manufacturing.

Analytical Procedures and Assays Are the Crux of Defining and Measuring CQAs
A CQA is a “physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality” (7). These attributes collectively define the safety, purity, potency, identity, and stability of products. However, identifying which properties are CQAs is not trivial. Usually, identification of CQAs is based on a priori knowledge of products and the MoA. That information is used to establish a broad range of potential CQAs, which can be assayed during process development (4). Such properties then are measured and their impact to product safety, and efficacy is determined by a formal risk assessment, in which a scoring system can be used to delineate noncritical and critical quality attributes. Even with process development and increased process knowledge and understanding, some CQAs might remain unknown because of the biological complexity of CGT products. However, lack of absolute knowledge of CQAs cannot be allowed to compromise CGT product safety, and a certain level of understanding is required to reach commercial production.

Measuring potential CQAs and determining how they are affected by changes in process parameters (or materials) are key requirements for establishing control over development processes. Therefore, adequate analytical procedures and assays must be identified and qualified for each potential CQA. Compendial tests are used where appropriate; however, fit-for-purpose analytical procedures must be developed for measuring many CQAs that are unique to each CGT product. In some cases, a single assay may not provide an adequate measure of a CQA, so it is more appropriate to use an assay matrix (8). CQAs are only as good as the analytical procedures used to define and measure them, so investing time and forethought into assay development for CQAs early in process development is important.

Analytical Procedures and Validation Processes Will Evolve During a Product Lifecycle
An analytical procedure has a similar method lifecycle to process validation and should be developed as early as possible to minimize risk. Analytical development should ensure that consistent methods are used throughout process development and manufacturing (9).

For process development, analytical procedures and assays may not necessarily be validated because many of the properties being measured may not constitute a CQA, or they do not directly correlate to a CQA. That is because during early process development, many product attributes are measured and then assessed to identify a smaller set of CQAs to focus on during later development stages, and only those CQAs will require validated assays. Also, assays can change significantly or be replaced by better assays as more process knowledge and understanding are acquired.
For new analytical methods or assays, it is important to evaluate the specificity, linearity, limit of detection (LOD), limits of quantitation (LOQ), range, accuracy, and precision early in product development (9). Methods also should be evaluated in terms of robustness to enable early adoption of those methods that are likely to be approved in late-phase clinical trials.

Identifying a design space early allows for an assay to remain relevant and minimizes the need for bridging studies between stages before the initial investigational new drug (IND) application to phase 3 trials. During IND submission, appropriate acceptance criteria for each CQA should be established based on data obtained by lots used in preclinical and/or clinical studies. Data from lots used for demonstrating manufacturing consistency and from stability studies as well as other relevant development data also can be used for determining such acceptance criteria (10).

Although validation of procedures usually is not required for phase 1, parameters to be assessed when moving into later phases can include accuracy, repeatability, precision, sample and reagent stability, LOD, LOQ, linearity, and range (9).

During early stages of product development, few specifications are finalized, and modifications can be made to preparations and dosages. So emphasis should be placed on the safety of subjects in proposed studies. Safety tests should discern the presence of extraneous material, adventitious agents, microbial contamination, and replication competent viruses. To maximize assay method sensitivity, an assay should be performed at the process stage at which contamination is most likely (10). Additional tests will depend on the type of therapy, but they should assess purity, toxicity, and stability as well as the physical, chemical, and biological characteristics of products to ensure that they meet acceptable limits of identity, potency, quality, and purity. In particular, assays used to determine dosage should be well qualified before clinical studies begin. Doing so will ensure comparability between doses from preclinical to clinical lots.

In addition to adhering to stringent validation requirements concerning repeatability and precision when moving into phase 3, it is important to understand the effect that changes in method parameters can have on analytical procedures. Hence, a systematic approach to assess method robustness (e.g., DoE) should be adopted. After initial risk assessment is performed, a multivariate experiment to understand factorial parameter effects on assay performance should be assessed. Factors that can be considered are changes in manufacturer and reagents as well as variability in time, temperature, and sample preparation. An investigation of specificity also should be conducted to validate identity, purity, and dosage tests. Procedures used to demonstrate specificity will depend on an assay’s objective. Qualification of all assays should be summarized in a report and in standard operating procedures (SOPs) along with batch records and analysis.

For commercial manufacturing, assays that measure the identity, strength, quality, purity, and potency of drug substances and drug products must be validated. Validation data should demonstrate that analytical procedures and assays are adequate for their intended purpose and conform to standards of required accuracy, sensitivity, specificity, and reproducibility. Suitable reference standards and materials should be identified and their use supported by documentation, including “qualification test reports and certificates of analysis (including stability protocols, reports, and relevant known impurity profile information) as applicable” (9). Validation data should be analyzed using sound and appropriate statistical procedures and evaluated against predetermined acceptance criteria.

A risk-based approach is used to determine whether manufacturing process changes necessitate revalidation of analytical procedures and methods during a product’s lifecycle. Even when a manufacturing process is unchanged, all validated and implemented analytical procedures must be monitored over a product’s lifecycle to ensure that it continues to be adequate for its intended purpose.

Evaluation of method performance should be conducted at regular intervals, with trend analysis performed to determine the need for reevaluation or optimization. It is important to note the significance of maintaining retention samples, if permitted by CGT product design, to allow for comparative studies when changes to analytical procedures are required.

Guidance documents from the FDA and other regulatory agencies on the setup of analytical procedures and process validation for products provide excellent starting points for defining and validating such products and their manufacturing processes. However, no one-size-fits-all recommendation exists in those guidelines because of the complexity of CGTs. Nevertheless, the analytical procedures used to qualify and validate products are key to getting approval. Below we discuss examples of how analytical procedures played important roles in demonstrating product quality and the effects on efficacy and safety.

Example 1: Identifying a CAR-T Phenotype Correlated to Efficacy and Safety
T cells have been genetically modified to express a CD19-specific CAR to treat B cell acute lymphoblastic leukemia (B-ALL) with promising results. Although it appears that robust proliferation of CAR T cells in patients prevents relapse, it has been challenging to define the CQAs that determine CAR T-cell expansion. That is because of the wide variability in doses, phenotypic composition, and therapies (chemotherapy and/or an allogeneic hematopoietic stem cell transplant as well as a lymphodepletion regimen) that patients were receiving before administration (11).

Although some immune stimulation and inflammation was expected, patient deaths because of severe cytokine release syndrome (CRS) and neurotoxicity after administration of CAR T cells emphasized the need for improvement and qualification. Potency assays for approved CAR-T products include measuring the percentage of CAR+ cells and the release of IFN-γ from product cells in vitro in response to CD19 stimulation. However, neither of those measurements has been correlated clearly to clinical efficacy.

Phenotypic assays have been conducted to characterize CAR T-cell products, but such assays have not been used to predefine the formulation of such products. Preclinical studies demonstrated the functional differences between CD4+ and CD8+ T-cell subsets. However, by using a phenotypic assay to measure a broad range of potential CQAs, it was discovered that central memory T cells (TCM cells) or naïve T cells (TN cells) demonstrated higher potency in eliminating CD19+ tumor cells in mice than did effector memory T cells (TEM cells) (12).

Heavily pretreated adult B-ALL patients show a wide variability in proportions of CD4+ and CD8+ T-cell subsets, which would make CAR T-cell products from those patients poorly defined. However, by controlling the ratio of CD4+:CD8+ CAR T-cells, results showed improved potency and predictability of the cell product (11). With a higher potency, the dosage can be de-escalated to reduce risk of severe cytokine release and neurotoxicity.

This case study highlights the normal progression of cell-based therapies, for which the complete mechanism of action is unknown. But as more knowledge is gathered about the product during clinical trial studies, its QTPP may change or evolve to include more CQAs that incorporate that new information. In this particular case, a new clinical trial was performed to test the effect of infusing a product with a predefined ratio of CD4+ and CD8+ T cells. Ultimately, the goal is to make CAR T-cell therapies more robust and effective.

Example 2: Developing Adequate Potency Assays as a Necessary Component of Comparability Studies
Potency is defined as “the specific ability of the product to effect a given result, as measured by appropriate laboratory tests (in vivo, in vitro or both), to confirm and compare the product quality, strength, and efficacy across different manufacturing lots during all phases of a clinical study” (8). Potency assays measure the therapeutic ability of products and include biological and/or analytical assays. Potency assays also help to define shelf life and enable equivalency testing.

Some challenges in establishing potency tests for CGTs include inherent variability of starting material, limited material for testing, short stability, lack of established reference standards, heterogenous product composition, incomplete understanding of MoA, and the in vivo fate of a product. Thus, it is unlikely that the most appropriate potency tests are established in the early stages of product testing. Such tests often evolve with deeper understanding of the product’s attributes and its mechanism of action.

Although the evolution of assays is necessary and justified, lack of uniform testing during product development poses the risk of inconsistency in interpreting results and inferring clinical outcomes. The most recent example of this is the case of Zolgensma (onasemnogene abeparvovec-xioi), which is an FDA-approved adenoassociated-virus–based gene therapy for pediatric patients under two years of age who suffer from spinal muscular atrophy (SMA) caused by biallelic mutations in SMN1. One potency test for that product includes in vitro quantitative and qualitative assays to estimate product concentration to determine dosage. During the biologics license application (BLA) review process, it was realized that one test used to determine dosage in phase 1 clinical trials needed to be replaced with a more precise and accurate potency assay, which was developed during later clinical phases. Although the product lots used in the phase 1 study were retested using the improved assays, parallel stability assays indicated that the product was unstable over long-term storage. Thus, retrospective analyses on older lots were unreliable. Although there is no dispute that the phase 1 clinical trials supported the therapeutic potential of this product for SMA, the results are not directly comparable with the phase 3 trial that used a different manufacturing process and improved potency assays.

The above example highlights the challenge of demonstrating equivalency between phases of a clinical trial when a potency assay may evolve or change between phases. The unknowns of the product’s stability made it difficult to compare lots.
Progressive potency assay development can be achieved by thorough product characterization and understanding of the CPPs and CQAs that affect biological activity, stability, and lot-to-lot consistency. Risk assessment during early stages could have identified stability as a CQA of high importance. Testing multiple potency assays during early product development helps in the building of an expansive collection of data for comparing and correlating suitability and in establishing the most appropriate assays for manufacturing and lot release of a specific CGT product.

Assays and the Future of CGT Manufacturing
As more CGT products are developed for clinical trials and commercialization, technologies for manufacturing and analytics will continue to evolve along with a deeper understanding of the biological mechanisms of such products. Convergence of both technologies and analytics will push the industry to adopt strategies such as real-time release testing and process analytical technology (PAT) to increase assurance of product quality, safety, and efficacy while reducing costs and shortening manufacturing times.

Working toward “off-the-shelf” allogeneic products for cell-based therapies also will reduce product variability by creating more streamlined industrial manufacturing processes and facilitating testing for lot-to-lot consistency. Analytical procedures and assays always will be at the forefront because they define products and demonstrate their safety and efficacy.

Working toward standardizing assays (through groups such as the National Institute of Standards and Technology, the Alliance for Regenerative Medicine, and the Standards Coordinating Body for Regenerative Medicine) and creating reference materials for the CGT industry also will be important to compare products across international manufacturing sites. Altogether, the goal is to create safe and effective products for global patient access.

References
1 CBER. Human Gene Therapy for Hemophilia Draft Guidance for Industry. US Food and Drug Administration: Rockville, MD, July 2018.

2 Stanton D. In the Pipeline: Surge of Cell and Gene Therapies Likely in 2020. BioProcess Int. 17(9) 2019; https://bioprocessintl.com/bioprocess-insider/therapeutic-class/in-the-pipeline-surge-of-cell-and-gene-therapies-likely-in-2020/

3 Hitchcock T. Manufacturing of AAV Vectors for Gene Therapy. Gen. Engin. Biotechnol. News. 37, 2017: 13.

4 Lipsitz YY, Timmins NE, Zandstra PW. Quality Cell Therapy Manufacturing By Design. Nature Biotechnol. 34(4) 2016: 393–400.

5 CDER/CVM/ORA. Guidance for Industry PAT: A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. US Food and Drug Administration: Rockville, MD, September 2004.

6 CBER/CDER/CVM. Guidance for Industry Process Validation: General Principles and Practices. US Food and Drug Administration: Rockville, MD, January 2011.

7 CBER/CDER. Guidance for Industry: Q8(R2) Pharmaceutical Development. US Food and Drug Administration, Rockville, MD, November 2009.

8 CBER. Guidance for Industry Potency Tests for Cellular and Gene Therapy Products. US Food and Drug Administration: Rockville, MD, January 2011.

9 CBER/CDER. Analytical Procedures and Methods Validation for Drugs and Biologics. US Food and Drug Administration: Rockville, MD, July 2015.

10 CBER. Draft Guidance for Industry: Chemistry, Manufacturing, and Control (CMC) Information for Human Gene Therapy Investigational New Drug Applications (INDs). US Food and Drug Administration: Rockville, MD, July 2018.

11 Turtle C, et al. CD19 CAR-T Cells of Defined CD4+:CD8+ Composition in Adult B Cell ALL Patients. J. Clin. Investigation 126(6) 2016: 2123–2138.

12 Sommermeyer D, et al. Chimeric Antigen Receptor-Modified T Cells Derived from Defined CD8+ and CD4+ Subsets Confer Superior Antitumor Reactivity In Vivo. Leukemia 30(2) 2016: 429–500.

Manoja Eswara is a development scientist, corresponding author Sarah Kwon (sarah.kwon.ccrm.ca) is a development scientist II, Steven Loo-Yong-Kee is a senior development associate, and Vanja Misic is a development scientist II, all at Centre for Commercialization of Regenerative Medicine, 661 University Ave #1002, Toronto, ON M5G 1M1, Canada.