PBMC-Based Framework Brings Standardized Validation to Flow Cytometry Assays in Cell Therapy Development

Cellular immunotherapies, from CAR-T cells to donor-derived immune products, promise highly personalized treatment, but their living, patient-specific nature makes them far harder to standardize than conventional pharmaceuticals. A team at Heidelberg University Hospital has now published a proof-of-principle framework, built on peripheral blood mononuclear cells (PBMCs), that shows how developers of cell therapies can validate their flow cytometry assays in a structured, fit-for-purpose way. The work, published in Frontiers in Medical Technology, addresses a persistent regulatory gap: while authorities demand validated analytical methods tied to a product’s mechanism of action, no widely applicable, stepwise validation framework has existed for the broader cellular immunotherapy field.

The researchers, led by Huixiu Han, Anita Schmitt, Michael Schmitt, and Lei Wang, used healthy-donor PBMCs as a model matrix and set out to quantify major leukocyte populations—T cells, B cells, monocytes, NK cells, and NKT cells—alongside early and late apoptotic cell fractions. Apoptosis was tracked with a fluorogenic Apotracker reagent and the vital dye 7-AAD, distinguishing early apoptotic (Apotracker-positive, 7-AAD-negative) from late apoptotic (double-positive) cells. The team stresses that the assay configuration was not developed as a product-specific release test; rather, it serves as a methodological template whose markers, acceptance criteria, and performance requirements must be tailored to each intended product and manufacturing stage.

Before any validation could begin, the panel itself had to be optimized. The researchers titrated key antibodies by evaluating both the staining index—the separation between positive and negative populations relative to the spread of the negative population—and the percentage of positive cells. For 7-AAD, volumes between 3 and 5 microliters performed best, with 4 microliters selected as the optimal balance; for Apotracker, 5 microliters delivered the highest staining index. Photomultiplier tube (PMT) voltages were then systematically mapped using eight-peak rainbow beads and unstained compensation beads, tracking the coefficient of variation, PMT linearity, signal-to-background ratio, and staining index across voltage settings. In the APC channel, dim peaks became indistinguishable at low voltages while bright peaks exceeded the dynamic range at high voltages; the optimal window emerged at 550 to 700 volts, with similar windows defined for FITC, PE, PE-Cy7, APC-Cy7, PerCP, and APC-R7 channels.

Panel robustness was further probed through steric hindrance and compensation checks. Fluorescence intensities from fully stained samples were compared with single-antibody controls, and most antibodies showed coefficients of variation below 10 percent. The exception was 7-AAD, whose variability—54.8 percent in median fluorescence intensity and 22.4 percent in percentage of positive cells—was attributed to the low frequency of apoptotic cells rather than true interference. Fluorescence Minus One (FMO) controls, in which each sample is stained without one specific antibody, confirmed minimal spectral spillover and validated the compensation matrix across all channels. A second staining panel reproduced these trends, supporting the general robustness of the approach.

With the panel optimized, the team moved to formal validation of specificity, sensitivity, linearity, precision, and robustness. Specificity testing across three healthy donors in two independent runs showed false-positive rates below 1 percent for CD14, CD3, CD45, and CD56, indicating minimal non-specific antibody binding. False apoptotic labeling rates remained below 10 percent for Apotracker and 7-AAD, meeting the predefined threshold, though the authors note that false positives rise as target populations become rarer—a reminder that higher acquisition thresholds are needed for rare-event analysis.

Sensitivity was established through dilution series in which stained cells were spiked into unstained samples. The lower limit of quantification (LLOQ) for CD3-positive T cells was a stained cell fraction of just 0.1 percent, corresponding to 32.3 detected events, with a coefficient of variation below the 30 percent criterion used for rare-event assays. Other populations showed LLOQs of 1 percent for monocytes (132.3 events) and B cells (17 events), 0.1 percent for NK cells (4.3 events), and 10 percent for NK-T cells (168.7 events). Limits of blank and detection were similarly low—for T cells, 0 and 3.4 events respectively—confirming the assay’s capacity to detect and quantify rare cell populations. Linearity, assessed by plotting measured cell counts normalized with counting beads against expected values on logarithmic scales, yielded R-squared values above 0.95 across all populations, with most cell types remaining linear at dilution ratios above 1:100 and B cells above 1:10. These results satisfy the criteria of the ICH Q2(R2) guideline on analytical validation.

Reproducibility testing spanned multiple dimensions. Intra-assay precision, evaluated by analyzing each sample three times within a single run, produced coefficients of variation below 10 percent for the main leukocyte populations and below 30 percent for rare apoptotic subsets. Comparing two instruments, a BD FACSCanto and a BD LSR II, gave comparable results for major populations (CV below 10 percent), though rare apoptotic events exceeded that threshold, reflecting differences in detector sensitivity, laser alignment, and calibration—an important caveat for laboratories switching to backup cytometers. Inter-operator reproducibility was strong: mean differences in cell frequency fell well below the 5 percent tolerance limit, and CV and relative change stayed under 10 percent for all subsets, underscoring the value of detailed standard operating procedures and thorough operator training. Median fluorescence intensity measurements for CD3, CD56, and CD66b showed statistically significant differences between operators, but these remained within the predefined tolerance in most cases.

Practical workflow considerations received equally rigorous attention. Stained samples held for 6 hours at room temperature or 24 hours at 4 degrees Celsius revealed that T cells and NK cells remained stable throughout, while B cells and NK-T cells drifted after 6 hours and monocytes after 24 hours; apoptotic cell fractions increased over time, particularly after 24 hours, arguing for immediate acquisition or validated storage conditions. A pre-mixed seven-color antibody cocktail stored at 4 degrees Celsius for 24 hours performed identically to a freshly prepared one, with CVs below 10 percent, supporting batch preparation to reduce pipetting errors. Lot-to-lot testing showed acceptable variability for major populations but greater variability for apoptotic cells, likely due to their low abundance and the use of non-GMP-grade reagents. Carryover testing—inserting buffer tubes between samples—revealed that one antibody lot exceeded the 1 percent acceptance threshold, with contamination above 5 percent for rare apoptotic monocytes and CD45-positive cells, prompting the recommendation of additional instrument rinsing cycles when carryover criteria are breached. Single versus double red-cell lysis produced statistically significant but practically minor differences, except for a notable shift in CD14 fluorescence intensity.

To anchor accuracy and longitudinal performance independently of biological donor variability, the team employed BD Multi-Check Control, a commercially stabilized whole-blood material with manufacturer-assigned, lot-specific target ranges. Across nine independent runs over four months, involving multiple operators and four reagent lots, all five monitored leukocyte populations remained within their assigned target ranges. The authors argue that incorporating such external controls should become a standard component of fit-for-purpose flow cytometry validation frameworks for cell therapies, enabling early detection of assay drift and providing traceable documentation for regulatory inspection readiness. The study was explicitly designed around the analytical lifecycle principles of ICH Q2(R2) and Q14, in which initial validation confirms suitability for intended use while continuous performance monitoring and risk-based revalidation follow as knowledge evolves.

The authors are careful about the boundaries of their claims. The framework supports structured analytical development and lifecycle monitoring, but applying it to a defined cell therapy product requires product-, matrix-, and decision-specific selection of measurands, markers, and acceptance criteria, followed by appropriate qualification or validation; the study itself does not establish regulatory compliance, GMP or GLP suitability, or clinical applicability. Rare-event measurements, they emphasize, demand particular caution, with sufficient gated-event numbers, population-specific sensitivity limits, and procedural controls to ensure reliability. Multicenter and inter-laboratory studies are now needed to assess broader transferability. Still, for a field where manufacturing consistency and regulatory compliance remain the chief obstacles to bringing living medicines to more patients, this PBMC-based template offers a concrete, technically grounded path forward—one that turns the vague regulatory mandate to validate analytical methods into an actionable, experimentally demonstrated workflow.

Subject of Research: Fit-for-purpose validation of flow cytometry assays for cell therapy development using a PBMC-based model

Article Title: A PBMC-based proof-of-principle framework for fit-for-purpose validation of flow cytometry assays in cell therapy development

Article References: Han, H., Schmitt, A., Hückelhoven-Krauss, A., Neuber, B., Tretter, T., Hinkelbein, M., Bauer, A.-K., Müller-Tidow, C., Schmitt, M., & Wang, L. (2026). A PBMC-based proof-of-principle framework for fit-for-purpose validation of flow cytometry assays in cell therapy development. Frontiers in Medical Technology, 8, Article 1905186. https://doi.org/10.3389/fmedt.2026.1905186

Image Credits: AI Generated

DOI: 10.3389/fmedt.2026.1905186

Keywords: flow cytometry, cell therapy, PBMC, analytical method validation, ATMP, apoptosis, CAR-T, ICH Q2(R2), potency assay, regulatory affairs, rare event detection, quality control