MSC manufacturing: cell banks, passages and potency assays

Where does MSC variability come from? Donor, tissue and serum

MSCs are adherent cells isolated from bone marrow, adipose tissue, umbilical cord and other sources. They can differentiate into bone, cartilage and fat, and they have properties tied to immune modulation and tissue repair.

Bone marrow is collected by aspiration, which is invasive, but it has the longest record in research and clinical use. Adipose tissue is easy to obtain in quantity, and its yield is said to be higher. Umbilical cord comes from tissue discarded at birth, and because the donor is young, proliferation tends to be higher. Even within one tissue, donor age, medical history and individual differences change the properties of the primary cells.

Cells made from the patient's own tissue (autologous) and cells made from a healthy donor and stocked in advance (allogeneic) handle variability differently (see autologous and allogeneic cell therapy (in Japanese)). In the allogeneic case, one donor yields doses for many patients. Temcell HS Inj. from JCR Pharmaceuticals was approved in September 2015 as the first allogeneic regenerative medical product in Japan (the PMD Act category for cell and gene therapy products). It uses MSCs isolated from the bone marrow of healthy adults and expanded. When the starting material cannot be uniform, donor selection and screening for infectious agents are where quality begins.

Even the enzymes that free cells from tissue vary from lot to lot. Some products, such as SERVA's Collagenase NB series, cover everything from research grade to GMP grade. For a clinical process, check early whether the dissociation enzyme comes in a GMP grade (see cell dissociation enzymes).

Serum is another source of variability. Fetal bovine serum was added to media for a long time, and three problems remain: lot-to-lot differences, concerns about infectious agents and residual xenogeneic components. That has strengthened the shift to serum-free, xeno-free media. Products such as FUJIFILM Irvine Scientific's PRIME-XV MSC medium and Sartorius's MSC NutriStem XF medium were made with the move to manufacturing in mind. Another approach uses human platelet lysate.

Going serum-free has a price. The adhesion support that serum provided is gone, so culture surfaces may need coating. The dissociation enzyme, the seeding density and the passage conditions all have to be re-established for each medium.

Re-establish the passage conditions whenever you change the medium, because swapping the medium alone gives cells that either do not expand or expand and change.

The ISCT criteria define what an MSC is, not whether it works

The minimum conditions for calling a cell an MSC are the three that ISCT (the International Society for Cell & Gene Therapy, formerly the International Society for Cellular Therapy) published in 2006, and they are widely referenced. The cells adhere to plastic. The surface markers CD105, CD73 and CD90 are 95% or more positive, while CD45, CD34, HLA-DR and others are 2% or less. And the cells can differentiate into bone, fat and cartilage. Markers are measured by flow cytometry, and differentiation is judged by staining after induction culture.

The ISCT criteria are a minimum, not a product specification. Meeting them with your markers means "may be called an MSC". It does not mean "works".

The potency assay, which measures whether the cells work, is still not standardized. Assays that use the ability to suppress T cell proliferation or the secretion of cytokines as the readout are under study. Because the mechanism of action has many sides, no single assay has been settled on. That gap shows in the Ryoncil history at the start of this guide.

The August 2023 FDA letter judged the evidence of efficacy in children insufficient. It also cited a lack of potency-assay evidence showing that the commercial product works consistently. According to Mesoblast, after further discussion with FDA it framed the remaining issue as the consistency of potency of the commercial product and said it would add potency assay data. Its resubmission in July 2024 included, in its account, additional data on the potency assay. FDA approved the product in December of that year. The indication is steroid-refractory acute graft-versus-host disease in children aged 2 months and older.

The lesson is that a potency assay is not something you build just before the marketing application. You prepare it at the stage of the product used in clinical trials, and you keep a record of how potency changes when manufacturing conditions change.

Build MSC quality control to show in data, batch after batch, that the defined properties are held, because raising strength is not the goal.

Passage limits and two-tier banks fix the cells' identity

The more you passage, the more the cells age. Growth slows, the cells become large and flat, and both differentiation ability and immune-modulating ability move. The longer you extend passages to gain cell numbers, the greater the risk to the properties you want to keep.

So you set an upper limit on usable passages. The basis for the limit is data you took on your own cells: at which passage the marker-positive rate falls below the criteria, and at which passage differentiation ability drops. Another company's numbers will not fit. You also have to define the starting point: counted from primary culture, or counted from the bank.

The starting point comes from a two-tier cell bank. The master cell bank (MCB) is the cells of a selected donor, expanded to be uniform and frozen in bulk. The working cell bank (WCB) is made by thawing part of the MCB and expanding it further, and it is the stock you draw on for each manufacturing run. You do not start from fresh donor tissue every time. You start from a bank whose characterization is done, and that gives lot-to-lot consistency (see MCB versus WCB (in Japanese) and cell banks (in Japanese)). Count passages from the MCB and record the passage number of each lot.

Freezing also needs careful design. The cells are protected with a cryopreservation medium containing DMSO, the freezing rate is controlled, and storage is in the vapor phase of liquid nitrogen. Some products, such as BioLife Solutions' CryoStor, come in versions that differ in DMSO concentration (see cryopreservation media). Whether the properties return after thawing is part of whether the bank is qualified.

A two-tier bank fixes donor-derived variability upstream and, at the same time, gives you a reference point for counting passages.

Scaling up with microcarriers means re-measuring the same attributes

To expand allogeneic cells to many patients' doses, piling up flat flasks hits limits in area, labor and aseptic handling. So the surface the cells adhere to is replaced with small spherical carriers called microcarriers. Put carriers such as Cytiva's Cytodex into a stirred tank and you fit the same culture area inside the tank (see microcarrier culture (in Japanese) and microcarriers).

The price is more control points: shear from stirring and aeration, cells moving between carriers, and harvest. At harvest the cells are detached from the carriers again with a dissociation enzyme, and cells and carriers are separated and collected. If the cells are damaged here, the work up to this point is wasted. Counterflow centrifugation equipment such as Sartorius's kSep is used for separation and washing (see washing and concentration for cell therapy (in Japanese)).

Conditions set in flat culture do not necessarily hold in a stirred tank. As with changing the medium, when you raise the scale you re-establish the conditions.

Do not judge microcarrier culture by the number of cells it produced. Check on the harvested cells whether they meet the ISCT criteria and still differentiate. The effects of shear and enzyme treatment appear in the properties, not in the count.

A passage limit set in flat culture does not carry over to a stirred tank until you measure the same attributes again on the harvested cells.

Eight checks for specifications, banks and contract manufacturers

Use these when you compare manufacturers or choose materials.

1. Does the specification include a potency assay in addition to surface markers and three-lineage differentiation? If not, what is measured instead, and how far has it been discussed with regulators?
2. Is there a history of manufacturing changes from the product used in clinical trials to the current product, with comparison data from before and after each change?
3. Are donor eligibility records, infectious agent test records, and records of tissue of origin and donor age in place?
4. Is the medium serum-containing or serum-free? If it contains serum, how are lots verified? When the medium was changed, were the passage conditions re-established and recorded?
5. Is the passage limit set from your own data, and is the starting point counted from the bank?
6. Are there characterization records for the MCB and WCB, and records of recovery after thawing?
7. After scale-up, are markers, differentiation ability and potency re-measured on the harvested cells?
8. Do the dissociation enzyme, the medium and the cryopreservation medium come in GMP grade, and does the contract make the supplier send you notice of changes?

For sales, starting by asking the customer "autologous or allogeneic, and are you at the stage of building a bank?" shows how much medium, enzyme and testing will be needed.

Materials and equipment can be compared on the directory pages for MSC culture media, cell dissociation enzymes, cryopreservation media, flow cytometry and MCB/WCB manufacturing and storage services. For washing and concentration equipment, the list of automated cell washing and concentration systems is a starting point.

Two other guides answer the next questions. How to measure whether cells work is in cell therapy potency (in Japanese), and manufacturing of the vesicles that MSCs release is in exosome and EV manufacturing (in Japanese).