Challenges of Cell Counting in Cell Therapy Products

Abstract

Cell counting is a common and fundamental cell measurement technique that plays a crucial role in the development and quality control of cell therapy products. However, accurate and reliable cell counting can be challenging owing to the complexity of cell preparations, diverse counting purposes, and various counting methods. This review summarizes the challenges encountered in cell counting for cell therapy products and provides strategies to improve the cell counting accuracy, thereby guiding the counting process and ensuring the quality of cell therapy products.

Graphical Abstract.

Introduction

Cellular therapy products represent novel treatment options that employ living cells to replace, enhance, or modify malfunctioning or absent cells in a patient, distinguishing them from small molecule drugs and biologics ¹ . Unlike these conventional therapeutic modalities, which are either synthesized chemically or sourced from living organisms and act by interacting with specific molecular targets or pathways or binding to cell surface receptors, cell therapy products leverage the inherent properties of living cells to elicit their therapeutic effects. These products can be derived from the patient’s cells (autologous), donors (allogeneic), or animal cells. Examples of cellular therapy products include cellular immunotherapies, cancer vaccines, and various types of stem cells. Excitingly, cell therapy products such as CAR-T (chimeric antigen receptor T cell), MSC (mesenchymal stem cell), HPC (hematopoietic progenitor cell), or TIL (tumor infiltrating lymphocyte) have garnered regulatory approval in key jurisdictions, such as the US, EU, China, and Japan, offering immense potential in fields such as regenerative medicine and oncology^2,3.

Cell counting, encompassing total and differential cell counting, is the cornerstone in the manufacturing and releasing of cell therapy products. The quantification of cells is typically expressed as cell concentration in suspension (number of cells per unit volume) or as the area density of cells attached to a surface (number of cells per unit area) ⁴ . Precise quantification of cell populations is vital for assessing the number, viability, and purity, essential for evaluating the potency and effectiveness of these therapeutic products. It serves as a vital tool for monitoring cell growth and replication throughout the manufacturing process and aiding the determination of the appropriate dosage for patient treatment.

The International Organization for Standardization (ISO) has issued two standards related to cell counting, namely “ISO20391-1:2018 Biotechnology—Cell Counting—Part 1: General Guidance on Cell Counting Methods” and “ISO 20391-2:2019 Biotechnology—Cell Counting—Part 2: Experimental Design and Statistical Analysis to Quantify Counting Method Performance.” Additionally, the development of “ISO/CD 8934 Biotechnology General Considerations and Requirements for Cell Viability Analytical Methods Part 1: Mammalian Cells” is currently underway. A collaborative workshop hosted by the National Institute of Standards and Technology (NIST) and the US Food and Drug Administration (FDA) in April 2017, titled “NIST-FDA Cell Counting Work-shop: Sharing Practices in Cell Counting Measurements,” further enhanced our understanding of cell-counting practices.

This study examines the intricate complexities of cell counting within the realm of cell therapy products, emphasizing challenges like cell heterogeneity, various counting objectives, methodologies, and factors that affect counting accuracy. These hurdles illustrate the intricate nature of conducting accurate cell counting and underscore the necessity of implementing strategies to enhance the precision of cell counting (as depicted in Fig. 1).

Figure 1.

A comprehensive overview of the complexities of cell counting, critical factors influencing cell counting, and strategies to improve accuracy.

Complexity of Cell Counting

Cell counting technology has a rich history dating back to the 18th century, with automation efforts already underway by the early 20th century. However, despite its advancements, a 2020 survey revealed that only 18% of respondents from key industries, such as biotechnology, pharmaceuticals, and manufacturing, expressed high confidence in their cell viability assay results ⁵ . This concern highlights the complexity inherent in cell counting and emphasizes the urgent need to bolster confidence in these assays.

At the heart of cell counting, intricacies lie in the composition and nature of the cell preparations. Typically, cell preparations include a diverse array of cell populations, potential contaminants, and the selection of a suitable suspension medium, all of which have a significant impact on the integrity and accuracy of the resultant cell counts ⁶ . The task of distinguishing living cell populations from their deceased counterparts represents one of the most daunting hurdles, given that cell therapy products are live entities, and their therapeutic efficacy is intrinsically linked to their viability ⁷ . Indeed, the demarcation between viable and non-viable cells is best conceived as a spectrum of cellular vitality, where such distinctions are inherently user-defined, contributing to a gamut of varying results when assessing factors such as membrane integrity, metabolic activity, molecular markers, and proliferative capacity ⁵ .

Moreover, the distinct characteristics of different cell types, such as size, shape, density, and propensity to form clumps, add further complexity to the counting process. For instance, MSCs typically exhibit a more substantial size than T cells and are marked by size heterogeneity. In contrast, human-induced pluripotent stem cells (hiPSCs) tend to aggregate, forming dense, clumpy structures that require entirely different counting approaches ⁸ .

Furthermore, in products containing a myriad of distinct cell populations, diversity is accentuated, with each cell type exhibiting varied properties and viability metrics. Disparities in viability among different cell populations have been observed following the thawing of cryopreserved peripheral blood stem cells (PBSCs) and peripheral blood mononuclear cells (PBMCs), encompassing various white blood cell subtypes ⁹ . Similarly, differences in viability have been documented within specific cell subpopulations, such as CD45+ and CD34+ cells in peripheral blood progenitor cells (PBPCs) ¹⁰ . The unique properties of each cell type make it challenging to develop a universally applicable counting method, necessitating tailored approaches to target desired cell populations.

Compounding this complexity, cellular debris often emerges as a common contaminant, encumbering cell counting endeavors. The presence of cellular debris can obscure accurate cell identification and lead to overestimation or underestimation of cell counts, thus necessitating debris removal techniques to ensure results. Besides, the choice of the suspension medium can pose its own set of challenges. Various media, such as culture medium, washing solution, and cryoprotectant can significantly impact the discernment of accurate cell counts. For instance, the presence of dimethyl sulfoxide (DMSO) complicates the fluorescence of acridine orange (AO), potentially leading to an underestimation of the cell quantity, particularly at higher DMSO concentrations ¹¹ . Additionally, salt solutions, such as saline, PBS, and DPBS, may influence the binding capacity of AO to DNA, thereby reducing the observable cell count ¹² . In our observations, employing PBS or sodium chloride injection can reduce the freshly-harvested T cell concentration by nearly 40%, principally owing to a decline in AO/PI-stained cells, a dilemma that can be remedied by opting for a culture medium as the suspension vehicle (Supplementary Figure S1). The presence of 2.5% to 10% DMSO has also been shown to diminish cell viability from approximately 95% to 90% in MSCs shortly after addition, possibly due to a decrease in AO/PI staining or an increase in cell death (Supplementary Figure S2).

In a broader context, the different objectives and methodologies employed in cell counting further speak of its complexity. Whether it is in-process estimations for feeding rates, passaging, and culture maintenance or the critical monitoring of cellular responses to external stimuli and the evaluation of cell viability as an indicator of therapeutic potency, every objective may demand distinct levels of precision and accuracy. The necessity for tailored counting protocols that cater to varying objectives amplifies the intricate nature of cell counting.

Critical Factors Influencing Cell Counting and Strategies to Improve Accuracy

For accurate cell measurements, precision, sensitivity, and stability are imperative in cell counting. This entails consistently hitting the bull’s eye, maintaining stability in repeated counts, and possessing ample resolution to discern minute changes. These prerequisites can be attained through a meticulous approach that encompasses the choice of an appropriate counting method, optimization of counting protocols, rigorous validation of methodologies, and establishment of thorough measurement assurance procedures.

Selection of Cell Counting Method

A diverse array of methods has shaped the landscape of cell counting techniques, each with distinctive utilities and constraints^13,14. The hemocytometer, a traditional approach, involves manually counting cells using a specialized chamber under a microscope ¹⁵ . This method, which was first developed specifically to obtain accurate cell counts, has remained a cornerstone for decades. However, this method is usually time-consuming and varies depending on the operator. Efforts have been made to enhance this method, such as improving the Neubauer counting chamber ¹⁶ . Furthermore, technological advancements have led to the development of automated systems, such as image analysis and impedance cell counters, offering a faster and more efficient means of analyzing numerous cell samples. Flow cytometry is a versatile and powerful technique that not only facilitates cell counting but also enables comprehensive analysis of cellular attributes such as size, granularity, and protein expression. The integration of internal calibration microspheres has bolstered the accuracy of flow cytometry in quantifying cell counts and determining cell viability ¹⁷ . Plating and colony-forming unit (CFU) counting, a common method, involves culturing cells on agar plates, allowing them to form colonies for manual or automated counting. Spectrophotometry indirectly measures cell density by assessing the intensity of light transmitted through a sample. DNA content quantification using fluorescence-based assays estimates cell numbers by measuring the fluorescence intensity of the DNA-binding labels. Additionally, metabolic activity assays, such as glucose, lactate, and ATP, can indirectly assess cell numbers based on cellular metabolic activity ⁴ . A detailed comparison of these methods is provided in Table 1.

Table 1.

A Detailed Comparison of Cell Counting Methods.

Cell counting method	Description	Advantages	Disadvantages
Hemocytometer	Typically consists of a glass slide with etched grids, containing tiny counting chambers for manual observation	• Low cost • Suitable for various types of cells • Visualization	• Manual operation • Highly susceptible to human error • Time-consuming
Automated image analysis	Utilizes optical, electronic, and image processing technologies to count and classify cells efficiently	• Fast speed • High throughput • High precision • Automated and easy-to-use	• Relatively high cost • Easily influenced by the types and conditions of samples and cells
Impedance cell counter	Using impedance technology, cells are counted and measured by detecting changes in electrical resistance as they pass through an electric field	• Fast speed • High throughput • High precision • Automated and easy-to-use	• Relatively high cost • Inability to differentiate between live and dead cells
Flow cytometer	By combining cell suspensions with fluorescent markers, it utilizes optical and electronic technologies to measure and analyze multiple parameters of individual cells	• High throughput • Multi-parameter analysis (size, shape, fluorescence intensity, etc.) • High sensitivity and accuracy	• High cost • Requires complex operation and technical experience
Plating and CFU counting	Cells are diluted and plated onto dishes, where each cell forms a colony, counted based on the number of colonies	• Low cost • Direct counting of live cells • High detection limit • Independence from instruments	• Low accuracy • Limited applicability • Long processing time
Spectrophotometry	Based on the ability of cells to absorb light at specific wavelengths, cell concentration is inferred by measuring absorbance	• Fast speed • High throughput • Wide applicability	• Inability to differentiate between live and dead cells • Susceptibility to sample interference • Need for standard samples with known cell concentrations
Cell number by DNA quantification	The cell number can be determined by DNA quantification, which involves measuring the amount of DNA present in a sample to estimate the number of cells	• High sensitivity • High throughput • Indirect measurement	• Susceptibility to various factors such as sample preprocessing
Metabolic activity assay	An indirect method that infers cell quantity by measuring the concentration of metabolites, such as glucose, lactate, ATP, etc., in the cell culture medium	• High sensitivity • High throughput • Indirect measurement	• Susceptibility to various factors such as cell type and culture conditions

Cell viability, along with viable cell counting, is paramount in cell therapy products and serves as a requisite release criterion for cellular products ¹⁸ . One commonly employed method involves assessing the membrane integrity using the trypan blue dye permeation assay ¹⁹ . Nucleic acid staining with fluorescent dyes such as acridine orange (AO), propidium iodide (PI), and DAPI (4′,6-diamidino-2-phenylindole) distinguishes between live and damaged cells based on membrane integrity ²⁰ . The LDH (lactate dehydrogenase) release assay measures the activity of enzymes in cell culture supernatants to indicate cell membrane damage ²¹ . Evaluating metabolic activity is vital, and methods such as the MTT (3-4,5-dimethylthiazol-2-yl-2,5 diphenyl tetrazolium bromide) assay and ATP (adenosine triphosphate) assay are commonly employed. The MTT assay detects reduced MTT compound production within cells ²² , whereas the ATP assay quantifies ATP levels in cells or culture supernatants to assess metabolic activity ²³ . Monitoring glucose consumption in culture media is also a conventional method for indirectly evaluating cell metabolic activity ²⁴ . Molecular markers such as Annexin V and PI staining help differentiate between apoptotic and necrotic cells based on phosphatidylserine binding and membrane integrity ²⁵ . Finally, the clonogenic assay assesses cell proliferation and clonogenic potential by observing the ability of individual cells to form colonies and sustain their proliferation ²⁶ . Other dyes, such as calcein-AM, 7-amino actinomycin D (7-AAD), Alamar Blue, and Neutral Red, as well as molecular markers, such as Caspases, TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling), and the detection of cell proliferation via BrdU/EdU incorporation during DNA synthesis, can also be utilized to assess cell viability^27–29. Notably, in some cases, such as those based on metabolic activity or cell proliferation, which is used also for counting cell numbers, only viable cells can be evaluated; thus, non-viable cells and total cells cannot be assessed. Kamiloglu et al ³⁰ . conducted a comparative analysis of cell viability assays, examining their underlying principles as well as their respective advantages and disadvantages.

It is widely acknowledged that different counting methods may yield varying cell counting and viability outcomes, consequently influencing the reliability of the cell counting results. A case study by Humpe et al ¹⁰ . demonstrated notable discrepancies in white blood cell (WBC) viability in peripheral blood progenitor cells (PBPCs) when comparing trypan blue exclusion to flow cytometry, highlighting the importance of method selection. Moreover, other research, such as Cai et al ⁹ ., evaluated the accuracy of different viability measurement methods, including the manual trypan blue exclusion assay, flow cytometry with 7-AAD or PI staining, automated cell counter with AO/PI or trypan blue staining, finding that manual trypan blue exclusion assays typically provided the most reliable results compared to automated and flow cytometry methods using different dyes. Conversely, studies by Ohayon et al ³¹ . and Radel et al ³² . found that automated cell counters excelled in reproducibility and precision, particularly when assessing human adipose tissue-derived mesenchymal stromal cells (AD-MSCs) and T cell release testing. Additionally, the choice of staining dyes also plays a critical role, as seen in Chan et al ³³ .’s report that trypan blue exclusion assays often show higher viability measurements than AO/PI fluorescence-based methods, primarily due to inaccuracies in dead cell count in the trypan blue method. Considering these findings, it is important to select appropriate counting methods and dyes designed for specific research objectives for accurate cell assessment.

When it comes to selecting a cell counting method, adhering to a “fit for purpose” principle is paramount^5,6. In the assessment of a cell-counting method, several critical considerations require attention. First is a comprehensive understanding of the sample, encompassing the identification of cell type, evaluation of impurities, and determination of the suspension medium. Crucial inquiries arise as follows: what are the characteristics of cell morphology and aggregation? What are the anticipated cell populations in the presence of multiple cell populations? What constitutes the suspension medium, and how does the sample stability within it fare? Therefore, the use of microscopy for sample observation and analysis is indispensable. Second, it is imperative to evaluate the selectivity of the counting method for the target cells. This involves understanding the underlying principles and their efficacy in distinguishing target cells within a suspension. Key questions arise: can it discern total cells, viable cells, non-viable cells, cell debris, and the desired cell population? Do impurities and suspension media affect the counting method? A comparison of the various methods and instruments is essential in this context. Finally, the tolerance for the counting results warrants consideration. Is it counting for process monitoring or product release purposes? What is the number of samples per measurement? Are there any Good Manufacturing Practice (GMP) requirements to adhere to? Collectively, these considerations ensure that the selected cell counting method aligns with the specific requirements of the intended application.

For instance, Yang et al.’s study on post-thaw viability of peripheral HPCs compared several methods, including membrane integrity assays, CD34+ cell enumeration, and CFU-GM clonogenic assay. Their findings suggest that the recovery of viable CD34+ cells and CFU-GM remained consistent after post-thaw incubation, indicating their suitability as viability assays for cryopreserved HPC intended for transplantation ³⁴ . Similarly, Chan et al ³⁵ . proposed a method using AO/PI image cytometry for measuring peripheral blood mononuclear cell (PBMC) concentrations, which demonstrated higher accuracy by eliminating counting errors induced by red blood cells (RBCs), platelets, and other debris. Huang et al. noted differences in viability measurements between AO/PI and AO/DAPI staining methods, attributing them to variations in the dead cell generation processes. Their research indicates that AO/PI staining exhibited robustness and sensitivity, especially for low-viability samples ³⁶ . These studies highlight the importance of selecting appropriate cell counting methods to distinguish desired cells.

Remarkably, cellular injuries may exhibit delayed manifestations, which can lead to an initial overestimation of viability ³⁷ . This issue is particularly problematic when viability assessments are conducted immediately after thawing, as such immediacy in evaluation can excessively inflate perceptions of cell recovery. It is critical to recognize that a significant proportion of cell death often becomes apparent only 24 to 48 h post-thaw ³⁸ . In a study by Kardorff et al., a comparative analysis of various cell viability assays was performed to examine their effectiveness on human mesenchymal stem cells (hMSCs) and human A549 lung cancer cells under different cryopreservation conditions. Their findings revealed that assays gauging membrane integrity often overestimate viability, possibly because of their inability to distinguish between cells with intact membranes and those fully capable of post-thaw survival and function. Conversely, metabolic activity assays such as Alamar Blue were highlighted for their superior sensitivity, precision, and reliability in accurately reflecting cell viability and proliferation post-cryopreservation, marking them as advantageous for a broad spectrum of formulations and cell types. The study underscored the efficacy of viability assays can vary significantly depending on cryopreservation formulations, cell types, and specific experimental contexts, thereby underscoring the need for thorough consideration when selecting the most appropriate viability assessment techniques ³⁹ .

Further extending the exploration to viability assessment methodologies, Pierce et al. adopted a combined experimental and computational approach aimed at identifying cell viability assays that could accurately predict shifts in cell proliferation. While their initial findings revealed a lack of strong predictive accuracy immediately after the assay (at 0 h), certain assays, specifically AO/DAPI, Annexin V/PI, and LDH, demonstrated significantly enhanced predictive accuracy at 24 h post-assay, suggest a temporal dimension to the assay’s predictive capabilities, a crucial aspect for accurate viability assessment ⁴⁰ .

Overall, these studies highlight the variance among various cell counting methods and stress the importance of selecting approaches customized to specific objectives, accounting for factors such as cell type, experimental conditions, and intended applications.

Processes Involved in Cell Counting

The process of cell counting involves several crucial stages, beginning with sample preparation. This encompasses a series of steps, including mixing and diluting the cells, washing, resuspending them properly, and finally, sampling and staining, all of which may introduce variability and play a vital role in ensuring the accuracy of the cell count. Particularly, the intricacies of mixing and sampling, which cover both the initial creation of the cell suspension and the post-sampling phase, are crucial. Improper handling during these phases can lead to unrepresentative samples, thereby skewing the accuracy of the cell count. In the sample preparation phase, challenges like the presence of cell clusters or aggregates (e.g., MSCs after Trypsin dissociation or iPSCs in culture) can lead to uneven sampling or cell damage if not adequately addressed.

It’s important to note that the variability in counting outcomes is easily influenced by differences in operational techniques among personnel, particularly for methods reliant on manual sampling or mixing. Additionally, the use and concentration of dyes, stains, or labels to distinguish cell types or viability can introduce variability. There’s also the risk of dyes adhering to tubing and instrument chambers, potentially leading to cross-contamination between samples if proper cleaning protocols aren’t followed. Extended exposure to Trypan Blue, known for its cytotoxic effects, can perturb both cell concentration and viability assessments, as documented by Strober ⁴¹ and Tsaousis et al ⁴² .

Furthermore, effective control over cell concentration and dilution is crucial to avoid underestimation pitfalls and ensure precision. The choice of diluent, such as phosphate-buffered saline (PBS), also merits attention owing to potential shear stress effects and prolonged incubation impacts, as elucidated by Chen et al ⁴³ .

Data collection and analysis yield a significant influence on the reliability of cell counting. With the increasing popularity of automated cell counters, the process has become increasingly streamlined. Notably, the precise configuration of parameters such as the minimum and maximum diameter, cell sharpness, cell circularity, focus, light intensity, exposure, and decluster degree remain crucial for accurate automated image analysis.

Combining this, conducting cause-effect analyses for each step, including sample, reagents, consumables, equipment, environment, and procedure, is instrumental in identifying key factors in the cell counting process^44,45. Building on this foundation, the application of Design of Experiments (DOE) not only aids in understanding the extent of the impact of these factors but also streamlines the process of optimal condition screening for cell counting^46,47. For instance, parameters such as sample hold times, dilution factors, mixing durations, staining procedures, and operator techniques are typically scrutinized through DOE based method, which systematically vary these factors to identify the most effective combinations and levels for accurate cell counting.

Ultimately, establishing a standardized and repeatable protocol for cell counting based on the preceding experiments is essential. A steadfast commitment is required to maintain the accuracy and consistency of every procedural step, thereby ensuring the reliability of the cell counting results.

Strategies to Improve Cell Counting Accuracy

Ensuring the accuracy of cell counting outcomes is crucial once an appropriate counting method and protocol have been established. To accomplish this, additional strategies are necessary, including validating cell counting methods and implementing rigorous process controls.

As outlined in the ICH Q2 (R2) guidelines, the validation process should cover various aspects including specificity, accuracy, precision, intermediate precision, linearity, range, detection limits, quantitation limits, and robustness ⁴⁸ . Notable examples of comprehensive validation methods and processes include the following. Daniela et al. performed a comparative validation of three trypan blue exclusion-based methods: manual, semi-automated, and fully automated. This evaluation involved assessing linearity and range using concentration standards, viability standards, and cells, with data plotted against expected concentrations using statistical software. Specificity was validated using specific matrices (medium or buffer used in sample preparation) to ensure discrimination of actual cells from similar particles that may interfere with accurate readings. Accuracy was determined by comparing the concentration and viability standards to theoretical values. Additionally, the precision (repeatability) was confirmed through six replicates of uniquely concentrated samples, suggesting that automated methods provide efficient alternatives for processing numerous samples ⁴⁹ .

Manzini et al. conducted a robust validation study on an automated cell counting method for human-induced pluripotent stem cells (hiPSCs). They utilized DPBS as the matrix for hiPSC samples to exclude potential contamination by particles, thus ensuring specificity. Linearity was evaluated using a series of 1:2 serial dilutions to obtain hiPSC samples, confirming linear regression and the proportional relationship between the measured and expected values. Comparative assessments of the automated and manual counting methods were used to determine accuracy, thereby confirming their consistency. Furthermore, precision was analyzed by examining both the repeatability and the intermediate precision of the automated method ⁵⁰ . The studies by Daniela et al. and Manzini et al. illustrate the effectiveness and superiority of automated cell counting methods over manual techniques, particularly in terms of specificity, accuracy, and precision.

Moreover, Dadgar et al. evaluated the accuracy, precision, and specificity of cell count and viability measurements for MSCs, T-cells, and iPSCs using two automated platforms (Vi-Cell and NC-200), and compared them with manual hemocytometer-based methods. The evaluation included accuracy assessments for each counting method and precision calculations of the in-house assay using percentage recovery rate replicates for each analyst and platform. Specificity was determined by the ability of each platform to differentiate between live and dead cells and to accurately report viability percentages. The findings indicated that NC-200 might be more suited for the production of T-cells, while Vi-Cell showed better performance with iPSCs, and both platforms were deemed suitable for MSC assessments ⁸ . These studies underscore the significance of selecting and validating precise cell-counting methods to ensure reliability and efficiency in biological research and clinical applications.

It is noteworthy that the complexity of cell counting and the lack of a universally accepted method or reliable reference material have made assessing the accuracy of cell counting challenging. The ISO 20391-2 guidelines provide a valuable framework for evaluating the quality of cell counting measurements using a dilution series experimental design. This design involves sample preparation through a dilution series, labeling of test samples, and measurement of cell counts, with statistical methods such as calculating average cell counts, measuring precision with the coefficient of variation (CV), and assessing fit with the coefficient of determination (R2) and proportionality index (PI) ⁵¹ .

In line with these guidelines, Sarkar et al. conducted a dilution series study with replicated samples and observations, employing human Mesenchymal Stem Cells (hMSCs) as a model to compare automated and manual cell counting methodologies. The findings from this study demonstrated that for the hMSCs in question, automated methods yielded superior results in terms of both precision and proportionality ⁵² . In a similar study by Richards et al., three cell counting techniques (Method 1: trypan blue; Method 2 and Method 3: AO/DAPI) were evaluated accuracy and consistency at various stages of cell therapy production, including Leukopak, Peripheral Blood Mononuclear Cells (PBMCs), T cell-positive selection, and T cell-negative selection. This study also uncovered significant variances in accuracy and consistency across different methods, underscoring the impact of methodological choice on cell counting outcomes ⁵³ . Overall, leveraging quality indicators from experimental design provides a robust framework for validating and monitoring the precision and proportionality of cell-counting methodologies, thus enhancing the validation and oversight of cell-counting practices.

Despite rigorous validation efforts, the measurement process in cell counting can be impacted by unanticipated daily variables. Challenges including incorrect focusing, the use of expired reagents, and inaccuracies in setting instrument parameters, can potentially skew results, underlining the critical need for robust in-process controls. To mitigate such risks, adopting measures like Installation Qualification (IQ), Operational Qualification (OQ), and Performance Qualification (PQ) of instruments at regular intervals is essential. Additionally, the use of reference beads, adoption of orthogonal methodologies, adherence to daily calibration routines, thorough data review, and the implementation of duplicate detection strategies are indispensable for maintaining accuracy ³⁸ . Moreover, compliance with standards like USP 1044, which recommends the use of at least two independent assays for assessing post-thaw viability, is crucial for ensuring methodological rigor ⁵⁴ . Peskin et al ⁵⁵ . introduced a method for establishing a reference focal plane using beads, specifically aimed at enhancing the precision of trypan blue-based viability measurements. Together, these strategies significantly contribute to the reliability and accuracy of cell counting procedures.

In conclusion, implementing validation processes, utilizing automated methods, and process control can significantly improve the accuracy and reliability of cell-counting measurements, ultimately enhancing the quality of research and therapeutic product development in cell biology.

Cell Counting Methods in Cell Therapy Products and Release Criterion of Marketed Products

In the dynamic field of cell therapy products, precise cell counting is essential throughout the therapeutic journey, from the acquisition of cells to their reinfusion into a patient, regardless of whether the cells are autologous or allogeneic. Cell counting of stem cell products, such as iPSCs, is performed at various stages, including blood sample collection, cell selection and expansion, reprogramming, colony picking and expansion, nuclease-mediated genome engineering, cell differentiation, cell purification, cryopreservation, and quality control (QC) tests ⁴⁶ . In the context of CAR-T therapy, cell counting occurs during patient leukapheresis, peripheral blood mononuclear cell (PBMC) isolation, T cell purification, T cell transduction, T cell expansion, product formulation, and QC tests.

Challenges in cell counting vary depending on the critical quality attributes associated with different cell therapy products ⁵⁶ . As for CAR-T products, ensuring purity by monitoring contaminants such as red blood cells, circulating malignant cells, and non-T cell leukocytes is critical. Additionally, the heterogeneity of MSC products, which can arise from different tissue sources and varying cell sizes within the same product, presents its own challenges. In contrast, accurately measuring cell aggregates is a significant concern for iPSCs. Moreover, in a GMP (Good Manufacturing Practice) setting, manual counting methods are prone to errors, making automation essential to reduce deviations. Besides, the importance of accurate cell counting also differs between in-process monitoring and release assays. Release assays require higher precision and can be more affected by the suspension medium, compared to in-process monitoring, where samples are typically resuspended in culture medium. Addressing these challenges requires an understanding of the specific variations of each product and the unique requirements of different manufacturing stages, thereby necessitating the selection of appropriate cell counting methods on a case-by-case basis.

To investigate the differences in cell counting methods for various products, we conducted a review focusing specifically on GMP-compliant cell production processes, analyzing studies from the past five years indexed in PubMed. A diversity of cell counting techniques have been extensively used in the evaluation of MSC products, including automated image-based cell counters, manual counting chambers, and flow cytometry. To distinguish viable from non-viable MSCs, several strategies are utilized, such as acridine orange/propidium iodide (AO/PI), trypan blue, and 7-AAD staining techniques^57–63. When assessing T and NK cells, the employed methodologies span automated image cell counters, manual counting chambers, automatic hematology analyzers, and flow cytometry. Viability is primarily assessed using trypan blue, AO/DAPI, DAPI, 7-AAD, and eFluor450 staining^64–78.

Additionally, a detailed search strategy was employed to investigate the differences in counting methods for various products. Based on the current literature, the widely utilized techniques in the manufacturing of cell therapy products are image analysis (both manual and automated), flow cytometry, and hematology analyzers. However, searches using broader terms such as “image analysis” fail to yield definitive statistical outcomes, while inquiries focused on “flow cytometry” risk overestimating statistical findings, as this technique is primarily employed for identifying expression markers rather than for cell counting. To our knowledge, commonly used dyes in image analysis counters include Trypan Blue, AO, PI, and DAPI, whereas cell counting via flow cytometry often involves PI, DAPI and 7-AAD. To address these considerations, a search strategy was adjusted, combining various cell therapy product terms with specific keywords such as “(‘Trypan Blue’) OR (‘AO’ AND ‘PI’) OR (‘AO’ AND ‘DAPI’) OR (‘acridine orange’ AND ‘propidium iodide’) OR (‘AO’ AND ‘propidium iodide’) OR (‘acridine orange’ AND ‘PI’),” “(Hematology analyzer),” and “(‘flow cytometry’ AND ‘viability’ AND [‘Propidium iodide’ OR ‘DAPI’ OR ‘7-AAD’]).” This strategy highlighted disparities in the volume of relevant publications concerning MSCs, T cells, NK cells, and HPCs, as shown in Fig. 2A. Although we searched for iPSC products, the yield was minimal, resulting in no analysis. Notably, image analysis is predominantly leveraged in the evaluation of MSC products, distinguishing them from other cell types. Cell counting via hematology analyzers appears to be more integral to the assessment of HPC products. Moreover, the range of counting methods applied to both T and NK cells showed analogous patterns, underscoring the strategic selection of counting methodologies tailored to each unique cell product.

Figure 2.

Comparison of cell counting and viability assessment methods across different cell therapy products. (A) The number of publications related to various cell counting methods, including automated image-based cell counters, manual counting chambers, flow cytometry, and hematology analyzers, for different cell therapy products (mesenchymal stem cells [MSCs], T cells, natural killer (NK) cells, and hematopoietic progenitor cells (HPCs)). (B) The number of publications related to different cell viability assessment dyes, including trypan blue, propidium iodide, DAPI, 7-AAD, acridine orange, and others, for the same cell therapy products.

As membrane integrity dyes are the primary method in the analysis of cell viability, such as Trypan Blue, propidium iodide, DAPI, 7-AAD, Calcein-AM, and Alamar Blue, we also analyzed this aspect and revealed disparities in the volume of relevant publications regarding MSCs, T cells, NK cells, and HPCs. Both T cells and NK cells demonstrated similar patterns, indicating a preference for certain dyes across these cell types. Trypan Blue was the most widely used dye across all cell types, followed by propidium iodide, especially in MSCs, T cells, and NK cells (Fig. 2B).

For cell therapy products, it is imperative to conduct release testing to evaluate both cell count/dose and viability, relying on precise cell counting assays. The FDA advocates the establishment of minimum cell viability criteria, which are generally set to above 70%. In addition, it is essential to define the optimal range of viable and functional cells to ensure therapeutic effectiveness^79,80. For instance, tisagenlecleucel [Kymriah®] specifies a minimum and maximum dose range of 0.2 to 5.0 × 10⁶ transduced viable T cells for individuals weighing ≤50 kg, and 0.1 to 2.5 × 10⁸ for those weighing >50 kg, with viability criteria set at 70%^81–83. Similarly, FDA-approved HPC products derived from cord blood require a total nucleated cell (TNC) viability exceeding 85% ⁸⁴ , and Prochymal®, a mesenchymal stem cell therapy, mandates a minimum of 70% viability ⁸⁵ . These standards play a pivotal role in ensuring that viability and cell count meet the specified release criteria, maintaining therapeutic integrity.

It is worth noting that deviations from these specifications, known as out-of-specification results, do occasionally occur ⁸⁶ . Research by Chong et al ⁸² . indicated that there is no significant correlation between CAR-T product viability, whether below 80% or at least 80%, and clinical outcomes for CTL019. However, in instances where achieving viability is challenging, supporting data should be provided to demonstrate that lower viability, alongside the presence of dead cells and debris, does not compromise product safety or efficacy during administration^80,87.

In summary, the selection of a cell counting method for commercialized cell therapy products is critically important, balancing precision, expediency, and regulatory conformity. The release criteria for these therapeutic products were designed to confirm their safety and effectiveness for clinical applications. Essential considerations include setting a minimum viability threshold, ensuring a precise number of cells per unit volume, confirming the specific cell type, and minimizing contaminants. Additionally, for certain cell types, functional assays may be employed to verify whether the cells maintain their therapeutic efficacy. Collectively, these measures safeguard the quality and therapeutic integrity of cell therapy products, thereby upholding their clinical and commercial application.

Summary and Future Perspectives

Cell counting is a fundamental analytical method in biomedical research and clinical diagnostics, with implications for the development of cell-based therapies, such as immunotherapies and stem cell treatments. It is crucial that cell counting methods are developed in coordination with product development from the preclinical to phase I to biologics license application (BLA) stages ⁸⁸ . The accuracy of cell counts in GMP settings is vital for guaranteeing therapeutic efficacy and safety. Cells, however, are not merely static entities; they are dynamic organisms. The success of cell therapy relies on leveraging the intrinsic or secretion-mediated bioactivities of live cells to mediate therapeutic effects, a characteristic that is unattainable through small molecules or conventional biologics. Therefore, cell counting becomes highly intricate and challenging. Efforts have been made to develop standards for cell counting to improve confidence in this critical process ⁸⁹ . The transition to automation over the decades has markedly enhanced the speed and feasibility of the processing of numerous samples. This review highlights the importance of a thorough understanding of cell characteristics, evaluation of the suitability of counting methods, detailed research and validation of the counting process and parameters, and implementation of quality control measures.

Looking ahead, the horizon for cell counting technologies appears promising, with opportunities to achieve rapidity, high throughput, and traceability on the rise. Innovations such as automated cell counters and flow cytometry are set to refine the precision and efficiency of cell counts further. Nitta et al ⁹⁰ . developed a rapid and high-throughput T cell immunophenotyping and viability assay using advanced automated cell counters. Blache et al ⁹¹ . have showcased advanced flow cytometry assays designed for the immune monitoring of CAR-T cell therapeutics, employing standardized 13-color/15-parameter assays to meticulously track CAR-T cells and monitor immune cell subpopulations. Barbau and Viey ⁹² presented in-line monitoring and cell counting of CAR-T cells in a production process based on an automated microscope. Additionally, Babakhanova et al ⁹³ . have introduced a quantitative, traceable method utilizing absorbance microscopy for the assessment of intracellular trypan blue uptake, thereby enhancing the discrimination between viable and non-viable cells. Microfluidic technology represents another leap forward, offering accurate and high-throughput methods to enumerate and characterize cells ⁹⁴ . Furthermore, the advent of image-based analysis systems coupled with sophisticated machine learning algorithms promotes the evolution of cell counting accuracy ⁹⁵ . Of note, integrating artificial intelligence (AI)—merging deep learning with traditional machine learning techniques—has emerged as a powerful tool for swift and precise cell detection and enumeration ⁹⁶ . Moreover, despite the in vitro nature of previous assays, the effective therapeutic use of cells often correlates with the survival of transplanted cells in vivo. The importance of tracking cell viability in vivo is underscored by unique challenges and recent developments in Quantitative NIR-II Fluorescence Imaging designed to address these issues^97,98.

In summary, the imperative for accurate and dependable cell counting in the realm of cell therapy is clear. Given the inherent heterogeneity of cells and the diversity of counting methods and objectives, the field is confronted with notable challenges. Nevertheless, through a concerted effort to understand and navigate these challenges, alongside the strategic application of appropriate methodologies, validation of counting processes, and standardization of protocols, researchers and developers are well-positioned to enhance cell counting accuracy, thereby securing the quality of cell-based therapeutic products. With the continuous advancement in cell counting technology, the prospect for more refined and efficient cell counting methodologies in cell therapy is both bright and attainable.