For iPSC Therapies, Manufacturing Consistency May Mean Preserving Function

Key Takeaways

• iPSC-derived cardiomyocyte products with similar identity, purity, and engraftment characteristics can still differ substantially in functional performance, highlighting the need to characterize what therapeutic cells can do as well as what they are.

• Cell-line selection and genetic engineering can influence downstream contractile, electrophysiological, and therapeutic properties, making the biological starting material an integral part of manufacturing strategy.

• Aggregate format, maturation state, and scalable suspension processing may affect product characteristics that extend beyond cell number and purity and therefore may need to be considered in manufacturing control strategies.

• Functional potency assays could help connect measurable in vitro properties with therapeutic performance, but the new NHP findings represent correlations rather than a validated predictive potency assay.

• Storage, hold time, and cryopreservation need to preserve functional properties as well as viability and identity, reinforcing the idea that manufacturing consistency for living therapies may ultimately mean preserving function from production through administration.

A highly purified cardiomyocyte product that survives transplantation and forms substantial cardiac tissue might appear to satisfy several important measures of manufacturing and biological success. A study of induced pluripotent stem cell (iPSC)-derived cardiomyocytes published in August 2026 in Nature Communications suggests that those measures can still leave therapeutic function unresolved.¹

In a non-human primate (NHP) model of myocardial infarction, investigators transplanted aggregates containing approximately 95–98% ventricular-like cardiomyocytes into infarcted cynomolgus monkey hearts. The aggregates averaged approximately 200 μm in diameter, and the animals received roughly 5–7 × 10⁷ cardiomyocytes per heart. Three months after transplantation, substantial human cardiac grafts were present.

The more consequential result emerged when the investigators compared products generated from two genetically engineered reporter iPSC lines. The lines, designated Amber and Ruby, were subclones of the same parental line but carried different reporter transgenes. Cardiomyocytes produced from both lines differentiated efficiently and generated substantial grafts. However, animals receiving Amber-derived aggregates showed substantial recovery of heart function, while no increase in left ventricular ejection fraction was observed in the animals treated with Ruby-derived aggregates in the line-specific comparison.

The difference could not be explained simply by successful versus unsuccessful engraftment. Human grafts formed from both cell lines, and no apparent line-related differences were observed in graft structure or cardiac marker expression. Functional testing, however, revealed substantial differences. Ruby-derived cardiomyocytes generated considerably less contractile force and displayed altered electrophysiological behavior, while transcriptional analysis identified differences involving cardiac contraction and conduction. The evidence linked those effects to overexpression of the red fluorescent reporter used in the Ruby cells.

That qualification is important. The study does not establish that ordinary clone-to-clone variation among unmodified iPSCs will necessarily produce similarly divergent therapeutic outcomes. The two lines had been genetically engineered, and the study specifically implicated effects associated with one reporter transgene. However, independent evidence shows that genetic variability among iPSC lines can affect molecular profiles, functional capacity, and differentiation potential.2

The larger manufacturing question follows directly. A process can repeatedly produce cells that satisfy expected measures of cardiomyocyte identity without guaranteeing that those cells possess equivalent therapeutic function. For living therapies, manufacturing consistency may therefore have to extend beyond producing the same type of cell to preserving the functions required for therapeutic activity.

For an iPSC-derived therapy, consequential sources of product variability can arise before differentiation begins.

The NHP study provides an unusually clear illustration. A modification introduced at the level of the reporter construct persisted through cardiac differentiation and aggregate manufacturing and was associated with differences in contraction, electrophysiology, transcriptional state, arrhythmogenicity, and in vivo recovery.¹

Cell line selection is therefore more than an upstream sourcing decision. Developers need to understand whether the selected line, subsequent engineering, and the clones chosen after engineering can reliably generate the functions required in the final product. A modification introduced for tracking, selection, safety, or another development purpose may have consequences that are not apparent from differentiation efficiency or familiar identity markers alone.

The broader biology of iPSCs supports that concern. Studies comparing iPSC lines have found that genetic background can contribute to donor-specific molecular characteristics and influence functional and differentiation capacity.² Those observations do not show that genetic variation will translate directly into clinically meaningful differences for every therapy. They establish that the starting cell population is biologically consequential starting material rather than an interchangeable substrate.

Selection of a production line may therefore need to account for more than expansion characteristics, stability, differentiation efficiency, and cellular composition. For a therapeutic product, developers may also need evidence that the line can reproducibly reach the functional state required at administration.

The same reasoning applies to genetic engineering. Confirming that an intended modification is present and expressed addresses only part of the question. The NHP results show how an engineered feature can alter downstream properties even when efficient differentiation and graft formation remain intact. A decision made early in development can propagate through the process and emerge later as a difference in therapeutic performance.

The contrast between the Amber and Ruby products exposes a central characterization problem.

Across production runs, the aggregate process generated cardiomyocyte populations with approximately 95–98% purity based on the cardiac markers used in the study, and both reporter lines showed similarly high expression of those markers.¹

Functional testing separated them more clearly. Bioartificial cardiac tissues prepared from the same batches used for transplantation generated markedly different contractile forces. The products also differed electrophysiologically and transcriptionally. In the animals, those differences aligned with divergent recovery and arrhythmia profiles.

For manufacturing, the distinction is between characterizing what cells are present and characterizing what those cells can do.

Identity, purity, viability, and dose remain indispensable. They simply do not capture every property that may be relevant for a therapy whose intended activity depends directly on cellular function. A lot composed almost entirely of cardiomyocytes may still differ from another cardiomyocyte lot in contraction, electrical behavior, or other functions that affect performance after administration.

The study does not establish which functional characteristics should become critical quality attributes for an iPSC-derived cardiomyocyte product, but it provides evidence that measurable in vitro differences can distinguish cell products that later behave differently in vivo.

Reproducible composition remains necessary. For therapies such as these, reproducible functional phenotype may prove equally important.

Maturation belongs in this discussion because cardiomyocyte function changes as the cells develop.

The suspension-derived cardiomyocytes used in the NHP study initially displayed an early ventricular-like electrophysiological phenotype. After transplantation, the grafts developed more organized myocardial characteristics, and gene-expression analyses indicated progressive cardiomyocyte maturation. The investigators also interpreted their results as suggesting that aggregate transplantation may support a higher degree of tissue maturation than previously described approaches using dissociated cells, although that comparison was made across studies and does not establish a direct causal advantage of aggregates.¹

The manufacturing issue is not maturation for its own sake. It is determining which functional state should be produced at the time of administration and how consistently the process reaches it.

If maturation influences contractility, electrical properties, or other functions relevant to therapeutic activity or arrhythmogenicity, it becomes part of product development. Culture conditions, differentiation timing, continued maintenance, and handling may all affect the functional state of the cells at administration.

More maturation cannot automatically be assumed to be better. The study does not define an optimal pre-transplant state. Instead, it shows why functional assays may ultimately be more useful than broad labels such as “mature” or “immature” when developers begin defining specifications and comparability strategies.

Physical product architecture adds another dimension to functional control.

The NHP study administered intact cardiomyocyte aggregates averaging approximately 200 μm in diameter rather than dissociating the cells before transplantation. Approximately 5–7 × 10⁷ cardiomyocytes per heart produced substantial grafts, and the investigators reported functional results comparable with previous NHP studies that used 10- to 20-fold higher numbers of dissociated human embryonic stem cell-derived cardiomyocytes.¹

That cross-study comparison does not establish that aggregate administration universally reduces the required therapeutic dose by that magnitude, but it highlights why product format deserves attention during process development.

Aggregate architecture creates manufacturing attributes that do not arise in the same way for a simple suspension of individual cells. Developers may need to control aggregate formation, size, uniformity, cellular composition, and physical integrity as part of producing a consistent product.

Scalable approaches for manufacturing this type of product are already being developed. A standardized stirred-suspension protocol has demonstrated aggregate generation with at least 100 million cardiomyocytes expected after 10 days of differentiation and was designed around process simplicity, robustness, and compatibility with good manufacturing practice.³  Separate work using stirred suspension systems across multiple iPSC lines has reported high cardiomyocyte purity and improved batch reproducibility compared with a standard monolayer differentiation approach, while identifying quality, inter-batch consistency, cryopreservation, and scale as continuing translational challenges.⁴

The physical state created during manufacturing therefore remains part of the product that reaches transplantation. Product architecture cannot be treated solely as a late formulation decision if it also influences how the cells are produced, maintained, and ultimately presented for administration.

The key development question is functional: whether aggregate characteristics influence survival, maturation, therapeutic activity, or other performance attributes strongly enough to warrant incorporation into the manufacturing control strategy.

Once cell line selection, maturation state, and product architecture can all influence functional phenotype, the practical question becomes how a manufacturer knows that a finished lot still possesses the biology the therapy is intended to deliver. That is the potency problem.

The NHP study provides a useful starting point because the investigators performed functional testing on material from the cardiomyocyte aggregate batches used for transplantation. Those cells were incorporated into bioartificial cardiac tissues, allowing contractile force to be measured. Amber-derived tissues generated considerably greater contractile force than Ruby-derived tissues, and those differences aligned with the subsequent divergence in cardiac recovery. Electrophysiological differences were also associated with different arrhythmia characteristics in vivo.¹

These observations do not constitute a validated predictive potency assay. The NHP groups were small, and additional work would be required to determine whether any particular measurement can reliably predict therapeutic performance. The experiment nevertheless moves the potency discussion beyond a theoretical concern. It demonstrates that a measurable functional property of a manufactured cardiomyocyte product can correlate with how that product behaves after transplantation.

The challenge is identifying which measurement, or combination of measurements, captures therapeutic activity closely enough to become useful for product control.

Contractility is an obvious candidate for cardiac replacement therapy, but the mechanism is not fully resolved. The study could not exclude contributions from paracrine or other indirect effects, and further work would be needed to define the role of electromechanical coupling with host myocardium.¹ A useful potency strategy therefore cannot be chosen solely because an assay is technically feasible or intuitively connected to the cell type. It needs a defensible relationship to the intended therapeutic effect.

The U.S. Food and Drug Administration’s (FDA) draft guidance on potency assurance for cellular and gene therapy products reinforces this broader view. The agency describes potency assurance as a strategy encompassing manufacturing-process design, process controls, material controls, in-process testing, and lot-release potency assays, with the objective of ensuring that released lots possess the specific ability or capacity to achieve their intended therapeutic effect.⁵ 

That framework fits particularly well with the lesson from the NHP study. The relevant functional difference appeared to originate far upstream at the level of the engineered line, persisted through differentiation and aggregate production, appeared in functional testing, and was subsequently associated with different in vivo performance.¹

A potency assay at release can reveal whether a lot meets a defined functional standard, but potency assurance begins earlier. If a particular functional attribute matters, developers need to identify the materials and process conditions that create or preserve it. Process characterization consequently becomes integral to the potency strategy, helping identify where relevant functions are established, altered, or lost.

This is where manufacturing consistency becomes more demanding for living therapies. A reproducible process has value because it repeatedly generates the functional properties the product requires. Functional assays provide a way to determine whether that process continues to do so.

The same functional standard has to survive the interval between manufacture and administration.

Cardiomyocyte aggregates have been maintained in stirred suspension culture for up to 35 days without a reduction in cardiomyocyte content, providing a means to separate production timing from downstream use.³ The aggregates used in the NHP study were also maintained in suspension for extended periods before transplantation.¹

That flexibility could be operationally valuable, but cardiomyocyte content alone cannot establish that every relevant product function remains unchanged throughout a prolonged hold period.

Cryopreservation illustrates the issue more clearly. An optimized method for cryopreserving iPSC-derived cardiomyocyte aggregates retained more than 90% expression of the evaluated cardiac markers, supported greater than 80% viable-cell recovery five days after thawing, and preserved spontaneous contraction.⁶

Separate work comparing fresh and cryopreserved iPSC-derived cardiomyocytes found measurable changes in morphology and contractile behavior after freezing and recovery. Cryopreserved cells showed reduced contraction amplitude and altered contraction timing compared with fresh cells at the reported time point.⁷

Those studies used different systems and do not demonstrate that cryopreservation compromises a therapeutic cardiomyocyte product. They do show why post-thaw viability and marker expression cannot automatically establish functional equivalence.

For a living therapy, hold time, freezing, thawing, recovery, and preparation for administration can potentially affect biological properties relevant to performance. Logistics therefore has to be qualified against the functional definition of the product rather than evaluated only by whether an acceptable number of cells survive.

The operational goal remains attractive: separate manufacturing schedules from procedure schedules and create a product that can be stored and delivered reliably. The development requirement is demonstrating that the logistical model preserves the attributes that made the lot acceptable in the first place.

These manufacturing questions are already moving into a clinical context. An allogeneic iPSC-derived cardiomyocyte spheroid product has entered phase I/II evaluation in patients with severe heart failure secondary to ischemic heart disease.⁸ That clinical product should not be equated with the experimental aggregates used in the NHP study, but its development illustrates that iPSC-derived cardiomyocyte products have progressed beyond preclinical experimentation.

Regulatory frameworks likewise place product understanding at the center of development. The FDA guidance for cellular therapies intended to treat cardiac disease addresses chemistry, manufacturing, and controls information together with preclinical and clinical development and includes the delivery system among relevant considerations.⁹ The FDA’s May 2026 final guidance on chemistry, manufacturing, and controls (CMC) flexibilities for human cellular and gene therapy products allows appropriate flexibility during development while preserving the need to satisfy applicable CMC requirements.¹⁰

The difficult question for developers is which aspects of a living product must remain consistent enough to support that development path.

The new NHP study offers one answer without pretending to provide the entire solution. High cardiomyocyte purity, appropriate marker expression, and substantial graft formation did not guarantee equivalent functional performance. Differences originating in the engineered cell line remained detectable in contractile and electrophysiological assays and were associated with differences in recovery and arrhythmogenicity after transplantation.¹

That observation connects what can otherwise look like separate manufacturing topics. Cell line selection establishes the biological starting point. Differentiation and culture determine the functional state produced by the process. Aggregate manufacturing defines the product’s physical architecture. Storage and handling must preserve relevant attributes through administration. Potency testing then provides a means to determine whether the final product retains the functions required for therapeutic activity.

Not every link in that chain has been clinically validated, and each therapy will require its own evidence connecting process variables, product attributes, and outcome. The value of the new study is that it shows why those connections need to be investigated.

For iPSC therapies, manufacturing consistency may ultimately require more than repeatedly producing the same cellular phenotype. The process must reproducibly create, preserve, and deliver the biological functions that make that phenotype therapeutic.

1. Gruh, Ina, et al. “Human iPSC-cardiomyocyte-aggregate cell therapy in non-human primates and correlation of heart recovery with contractile and electrophysiological cardiomyocyte properties.” Nature Communications. 17: 9066 (2026).

2. Kyttälä, Aija, et al. “Genetic Variability Overrides the Impact of Parental Cell Type and Determines iPSC Differentiation Potential.” Stem Cell Reports. 6: 200–212 (2016).

3. Kriedemann, Nils, et al. “Standardized production of hPSC-derived cardiomyocyte aggregates in stirred spinner flasks.” Nature Protocols. 19: 1911–1939 (2024).

4. Prondzynski, Maksymilian, et al. “Efficient and reproducible generation of human iPSC-derived cardiomyocytes and cardiac organoids in stirred suspension systems.” Nature Communications. 15: 5929 (2024).

5. Potency Assurance for Cellular and Gene Therapy Products: Draft Guidance for Industry. U.S. Food and Drug Administration. Dec. 2023.

6. Becker, Fabienne, et al. “Improved cryopreservation of cardiomyocyte aggregates differentiated from GMP iPSC in a 3D culture format.” Scientific Reports. 16: 1640 (2026).

7. Kowalski, Kathrin, et al. “Cryopreservation alters contractile function of human induced pluripotent stem cell-derived cardiomyocytes.” Scientific Reports. 16: 20353 (2026).

8. “A Study of iPS Cell-derived Cardiomyocyte Spheroids (HS-001) in Patients With Heart Failure (LAPiS Study).” ClinicalTrials.gov. NCT04945018. Last updated 22 Sep. 2025.

9. Cellular Therapy for Cardiac Disease: Guidance for Industry. U.S. Food and Drug Administration. Oct. 2010.

10. Chemistry, Manufacturing, and Controls Flexibilities for Developing Human Cellular and Gene Therapy Products for a Biologics License Application: Guidance for Industry. U.S. Food and Drug Administration. 5 May 2026.