The goal of cell therapy developers is to produce safe, efficacious, and consistent products that help patients—many with intractable diseases. Analytical methods are critical to achieving this aim and to the overall development process. These methods form the basis under which all critical development decisions are made, including what manufacturing equipment to use, which genetic engineering methods produce the best combination of performance and viability, what media system and feeding schedules perform the best, and what patient populations to target.
Cell therapy manufacturers face several challenges in developing the appropriate tests to analyze and accurately characterize their biological products. This section will discuss these challenges, the current characterization strategies and tools that manufacturers use, and some future trends to overcome existing hurdles.
Defining product specifications
As defined in ICH Q6B, Guidance on Specifications:Test Procedures and Acceptance Criteria for Biotechnological/Biological Products, characterization of a biological product “includes determination of physiochemical properties, biological activity, immunochemical properties, purity and impurities” . A comprehensive understanding of the biological product allows the appropriate specifications to be established. At a higher level, each specification contributes to a set of “criteria to which a new drug substance or new drug product should conform to be considered acceptable for its intended use” . By definition, a specification consists of the assay or test, the protocol for performing it, and the numerical limits, ranges, or other observations that define a product’s acceptance criteria. These acceptance criteria can be divided into several categories—Identity, Purity, Potency, Safety, and Other—which are defined in Table 1.
Table 1. Categories of acceptance criteria for new cell therapies.
While the concepts of full product characterization and release specifications are easy to comprehend, in practice they are very challenging to implement.
Challenges in characterizing cell therapies
In 1982, a new era of pharmaceuticals began with approval of the very first biologic—recombinant insulin. In the time since, more complex recombinant proteins and monoclonal antibody therapeutics have been developed, many of which have become best-selling drugs. While seen as more challenging to develop, manufacture, and distribute when compared to small-molecule medicines, these therapeutic modalities pale in comparison to the complexities associated with “living” cell therapy development, such as CAR T cell therapy.
To begin, human cells are large—each consisting of millions of protein molecules. Human cells also have a complex structure consisting of a membrane, cytoskeleton, and organelles that carry out specific functions. In addition to their large size and diverse biochemical makeup, cell therapy products are heterogenous and dynamic, continually interacting with and responding to their environment. This dynamism makes it impossible to fully characterize all properties and functions of a heterogenous mixture of cell types that could change depending on how they are manufactured, stored, and administered. Each characterization assay only reveals a single or limited number of attributes at a point in time. The selection of the most appropriate cellular attributes to test coupled with the inherent limitations of each assay method produce a formidable challenge that the industry will need to address to realize the field of cell therapy’s full potential.
Nearly all cell therapies start with an initial cell source derived from donor tissue. Donor-to-donor variability presents a tremendous challenge when trying to achieve a consistent, predictable manufacturing output. For autologous therapies, additional variability occurs because patients present varying severity of illness and/or have undergone several treatments prior to providing tissue. These factors make it challenging to set meaningful specifications that ensure consistent production of high-quality therapeutic products.
Quality by design approach
More recently, cell therapy developers have been implementing Quality by Design (QbD) principles in their product development process with the goal of providing the highest quality products to patients. The concept was first described by engineer Joseph M. Juran , and later the FDA highlighted its utility in pharmaceutical development in their 2007 report, “Pharmaceutical Quality for the 21st Century: A Risk-Based Approach” . The overriding principle of QbD is “quality should be built into a product” via “a thorough understanding of the product and process”  and process control. The QbD systematic approach starts with predefined objectives and is based in sound science. Moreover, this approach seeks to understand the risks “involved in manufacturing the product and how best to mitigate those risks .”
As an example, the FDA presented a general QbD scheme  (see Figure 1). Generally, the QbD process starts by defining the end goal or product, with developers formulating a hypothesis on a product’s Mechanism of Action (MoA). The MoA describes the specific action(s) a cell product will produce to achieve a desired therapeutic effect. The MoA then informs the Target Product Profile (TPP), which describes the desired attributes of a product, including safety and efficacy-related characteristics. Next, the developer must establish which critical properties or Critical Quality Attributes (CQAs) must be controlled to achieve the desired clinical outcome. CQAs as defined are defined as “a physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to assure the desired product quality . A CQA is identified by the risk of harm to a patient if that CQA is not met. In cell therapy, the “product is the process” is commonly heard; the direct corollary in QbD are Critical Process Parameters (CPPs), which are a key part of a manufacturing control strategy. A CPP, per ICHQ8(R2), is “a process parameter whose variability has an impact on a critical quality attribute and therefore should be monitored or controlled to ensure the process produces the desired quality .”
Figure 1. Quality by Design (QbD) principles. This general scheme provides a decision framework that cell therapy developers must make to define and control the quality of a cell therapeutic. Based on a scheme presented by Dr. Finn from the FDA .
According to the FDA presentation, “CQAs and CPPs are used together to ensure <product> quality and manufacturing consistency” . Specific examples are provided in Table 2.
Table 2. Examples of CQAs and CPPs  .
|Critical Quality Attributes||Critical Process Parameters|
|Acceptance criteria for source material||Action limits for specific steps|
|Criteria for intermediates||Equipment performance|
|In-process and release criteria||Process limits|
The CQAs then become a product’s in-process and release specifications, with the latter determining a product’s suitability for its intended use (i.e., to treat a patient). However, these criteria are not fixed and are continually evaluated in an iterative process. Over time, a developer will gain additional product characterization data, accumulate more insight into a product during clinical trials, and gain more experience manufacturing the product. Such new data and its findings can justify revising, adding, or removing a specification (Figure 2).
Figure 2. Systematic and iterative approach to identifying CQAs that are clinically relevant. As cell therapy manufacturers gain more information on their product from a variety of sources, they will revise and/or add specifications.
In summary, the QbD approach is a way to link the patient, therapeutic product, and manufacturing process together by monitoring clinical outcome (safety and efficacy), CQAs, and CPPs. In order for the approach to work, developers must choose fit-for-purpose assays that assess the essential attributes that predict a product’s quality with appropriate performance.
Regulatory perspective and development stages
While investment in cell therapy development has grown exponentially recently, it is still a young industry. Consequently, the regulatory landscape is evolving as the industry and regulators learn more about how the various types of living cell products are manufactured and how they behave when administered to patients. Given the complexity of these products and the diversity of indications they are meant to treat, no single framework exists to govern how these therapies are evaluated. Instead, each characterization scheme must be tailored to a particular product. Several guidelines are available for characterization and analytical methods, but only offer overarching principles and recommendations to help developers (Table 3).
Table 3. Examples of assay measurements and platforms.
Flow cytometry, PCR, methylation-PCR, microarray
|Authentication||Short Tandem Repeat (STR), HLA-PCR or NGS||✓||✓|
|Purity 21CFR600.3||Impurity profile (Cas9, host cell DNA, host cell protein, residual vector DNA )|
|Raw material residual (e.g., Benzonase, BSA, antibiotic resistance gene, transfection reagent, column leachables etc.)||qPCR, ELISA, HPLC, mass spectrometry||✓|
|Residual bead||Flow cytometry, cell counting||✓|
|Ratio/Empty full particles||qPCR/ELISA, HPLC, electron microscopy, ultracentrifugation||✓|
|Contaminating cells||Flow, qPCR, sequencing||✓||✓|
|Potency 21CFR600.3||Functional assay|
|Surrogate assay||Cell type-specific||✓||✓||✓|
Cell-based assay, qPCR
|Genomic stability||Cell-based assay, microarray, NGS||✓|
|Vector aggregates||Dynamic Light Scattering (DLS)||✓|
|Sterility & Adventitious agents 21CFR610.12 USP<71>||Sterility, endotoxin, mycoplasma, virus|
Growth promotion test/bacteriostatis & fungistasis tests, LAL, PCR
|Vector genome titer (VG)|
|Infectious genome titer (IG)|
*Master Cell Bank
In general, specific product assays are evaluated based on whether they are fit-for-purpose and performance. Developers generally take an incremental approach (Figure 3). For early-stage trials, regulators appear more interested in the rationale for the developer’s choices, specifically demonstrating suitability of the method, justifying the choice of an attribute (CQA) to measure, and providing evidence and interpretation of results over the entirety of a product’s developmental history. In these early stages, acceptance limits may have wider ranges. In later stages, the assays and acceptance criteria become more defined and the focus turns to method validation and a statistical approach to method capability. Table 4 describes the eight essential steps of method validation—Accuracy, Precision, Specificity, Detection limit, Limit of quantitation, Linearity, Range, and Robustness—as outlined in ICH Q2/R1 Validation of Analytical Procedures .
Table 4. Method validation steps of an analytical procedure as defined in ICH Q2/R1 Validation of Analytical Procedures .
“The accuracy of an analytical procedure expresses the closeness of agreement between the value which is accepted either as a conventional true value or an accepted reference value and the value found.”
“The precision of an analytical procedure expresses the closeness of agreement (degree of scatter) between a series of measurements obtained from multiple sampling of the same homogeneous sample under the prescribed conditions.”
|Specificity||“Specificity is the ability to assess unequivocally the analyte in the presence of components which may be expected to be present.”|
“The detection limit of an individual analytical procedure is the lowest amount of analyte in a sample which can be detected but not necessarily quantitated as an exact value.”
|Limit of Quantitation|
“The quantitation limit of an individual analytical procedure is the lowest amount of analyte in a sample which can be quantitatively determined with suitable precision and accuracy.”
“The linearity of an analytical procedure is its ability (within a given range) to obtain test results which are directly proportional to the concentration (amount) of analyte in the sample.”
“The range of an analytical procedure is the interval between the upper and lower concentration (amounts) of analyte in the sample (including these concentrations) for which it has been demonstrated that the analytical procedure has a suitable level of precision, accuracy and linearity.”
“The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small, but deliberate variations in method parameters and provides an indication of its reliability during normal usage.”
Trends and future
Most cell therapy developers will affirm that product characterization and analytics are some of the biggest challenges facing the industry. Until progress is made on this front, cell therapies will be relegated to a last line of therapy and never reach their full potential. What then are some trends, opportunities for improvement, and where the field is headed in the future?
Flow cytometry has become an indispensable tool in cell biology, immunology, and cell therapy given its versatility. With a single platform, a developer can perform a wide variety of tasks, including for example, cell counts, cell viability tests, transduction/genome editing efficiency assays, and phenotyping. Additionally, flow cytometry can quantify the subpopulations of cells with a particular set of attributes, which is especially important in CAR T therapy. Despite its utility in cell therapy research, flow cytometry is not easily transferable into the GMP manufacturing environment primarily because of the variability inherent in the method. Multiple sources contribute to this variability, including operator, reagents, instrument setup, sample preparation, and data acquisition and analysis. However, flow cytometry will never be completely replaced, challenging suppliers to introduce solutions to improve its compatibility with GMP environment (e.g., utilizing automated flow cytometry gating for analysis).
Wherever possible, the cell therapy field is moving away from cell-based methods to molecular-based methods because the variability of human cells makes cell-based methods more difficult to implement. Molecular methods offer numerous advantages over cell-based ones. First, molecular methods tend be to more sensitive. This sensitivity is important for such cases as identifying potentially harmful contaminating cell types such as undifferentiated induced pluripotent stem cells in a differentiated cell product or contaminating B cells in a CAR T product for B cell malignancies. Next, molecular assays are more readily standardized, ensuring that an assay run on different instruments, by different operators, at different sites, and at different times yields consistent results. Last, molecular assays typically require less material for the test. Some autologous cell therapy developers use up 50% of the dose for quality control testing. Under these conditions, if patient’s starting material is limited, a developer may not have enough cells for a full therapeutic dose. In the future, developers will migrate to assay platforms that require less input material, most of which would be molecular based.
Single-cell analysis is becoming increasingly prevalent and important in cell therapy, especially after recent events with lentivirus-transduced stem cell products in clinical trials [7,8] .This event highlighted the need to understand exactly where transgenes integrate in the genome and if possible, determine if any endogenous genes have been disrupted. Overall, the industry is pivoting towards gene-modified cell therapies (e.g., CAR T), and as a result, the FDA is requesting that developers identify the location of a transgene on a single-cell level. With the advent of genome editing technologies and ever more complex engineering steps, it will also be important to quantify efficiency and demonstrate specificity and safety of the edits.
Many clinical and commercial stage developers are working next-generation manufacturing processes to shorten the duration of manufacturing to decrease overall labor and facility costs and improve throughput. Some studies have suggested that in-process and release testing contribute up 25–30% of production costs. These high costs are particularly burdensome in autologous cell therapy, because the entire cost of manufacture is allocated to a single drug for a single patient. The ability to multiplex compatible assays is one way to reduce these testing costs. Likewise, more rapid testing methods will be key to shortening vein-to-vein time. For example, commercially available rapid qPCR-based mycoplasma assays could be used in place of the 28-day USP <63> culture method. Another area of opportunity would be rapid sterility testing that several groups are working on, including Standards Coordinating Body.
Over the last decade, several companies have partnered on applying digital technology solutions to healthcare manufacturing [9,10]. Scientists from the National Institutes of Health and National Institutes of Standards and Technology published a groundbreaking paper on the implementation of artificial intelligence (AI)-based quality control of their stem cell-derived product for treatment of age-related macular degeneration. The algorithm they developed analyzes images obtained by quantitative bright-field absorbance microscopy and are able to determine maturation level of retinal pigmented epithelial cells. Such AI-enabled, image-based quality control methods allow for non-destructive monitoring of cell products and could enable adaptive manufacturing and real-time release. With digital methods like these, the potential exists to revolutionize the way cell therapeutics are tested and released, improving manufacturing throughput and reducing the time for delivery to patients.
As the field starts developing commercial-ready manufacturing processes, closed automated manufacturing systems will be deployed. Key to the implementation of these systems will be Process Analytical Technology (PAT). The FDA defines PAT as a “system for designing, analyzing, and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes” . The ability to make real-time, sensor-based measurements of critical process parameters, such as metabolite production or nutrient consumption, could improve a developer’s understanding and allow finer control of the manufacturing process. While PAT has been successfully implemented in large molecule biologics production, it is still in the early days of implementation in the cell therapy field. PAT designed specifically for cell therapy will be key to developing robust manufacturing operations that provide a steady supply of life-changing medicines to patients in need.
- ICH Q6B-Guidance on Specifications: Test Procedures and Acceptance Criteria for Biotechnological/Biological Products. European Medical Agency.
- Juran JM (1992) Juran on quality by design: the new steps for planning quality into goods and services. Free Press.
- Pharmaceutical Quality for the 21st Century: A Risk-Based Approach. (2007) FDA.
- Finn, F. (2018) Early-Stage Manufacturing Considerations for Cell Therapy Products. Presentation at CASSS Cell & Gene Therapy Products.
- ICH Q8(R2) Pharmaceutical Development. European Medical Agency.
- ICH Q2/R1 Validation of Analytical Procedures. International Conference on Harmonization.
- bluebird bio Provides Updated Findings from Reported Case of Acute Myeloid Leukemia (AML) in LentiGlobin for Sickle Cell Disease (SCD) Gene Therapy Program. (2021) bluebird bio.
- bluebird bio Provides Update on Severe Genetic Disease Programs and Business Operations. (2021) bluebird bio.
- Oxford Biomedica announces R&D collaboration with Microsoft to improve gene and cell therapy manufacturing using the intelligent cloud and machine learning. (2019) Oxford Biomedica.
- AWS Announces Strategic Collaboration with Novartis to Accelerate Digital Transformation of Its Business Operations. (2019) BusinessWire.
- Guidance for Industry PAT—A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. (2004) FDA.
Intended use of the products mentioned in these documents vary. For specific intended use statements, please refer to the Instructions for Use (IFU).