How Are Stem Cells Actually Manufactured? Inside the Lab Process

Clinic marketing pages talk about "your own stem cells" as if the process ends at the needle. It doesn't. Between tissue collection and the syringe sits a manufacturing pipeline — sourcing, isolation, banking, expansion, testing, freezing, and shipping — and how carefully each step is executed is arguably the single biggest driver of whether a cell product is safe and consistent. This article walks through that pipeline stage by stage, and explains the regulatory line (21 CFR Part 1271, and the much smaller category of FDA-licensed products) that separates rigorously manufactured cell therapies from the investigational and unapproved products sold at many clinics.

Where the Starting Material Comes From

Most cell products used in U.S. clinics originate from one of three tissues: umbilical cord tissue (specifically the gelatinous Wharton's jelly connective tissue surrounding the cord vessels), bone marrow aspirate, or adipose (fat) tissue removed by mini-liposuction. Cord tissue is typically collected, with consent, after a scheduled birth and would otherwise be discarded; donors go through infectious-disease screening and eligibility determinations required under FDA's donor-eligibility rules. Bone marrow and adipose tissue, by contrast, are usually collected directly from the patient who will receive the product (autologous use), which changes some of the regulatory calculus but not the need for careful, sterile collection technique.

Isolation and Cell Banking

Raw tissue isn't a drug product — cells first have to be separated out. Wharton's jelly is typically minced and digested (enzymatically or via an "explant" outgrowth method) to release mesenchymal stromal cells; adipose tissue undergoes collagenase digestion to yield the stromal vascular fraction, from which stem cells are cultured out; bone marrow is processed by density-gradient separation or plastic adherence. In rigorous manufacturing, isolated cells from a single qualified donor lot are expanded once into a master cell bank, which is characterized in depth and then used to generate smaller working cell banks — the tiered system (similar to ICH Q5D banking practice used in biologics) that lets a manufacturer produce many consistent batches from one thoroughly vetted starting stock, rather than starting from scratch, and with unknown variability, each time.

What a GMP Facility Actually Requires

FDA's current Good Tissue Practice (cGTP) regulations, under 21 CFR Part 1271 Subpart D, require manufacturers to operate facilities "of suitable size, construction, and location to prevent contamination," with defined work areas, environmental monitoring (temperature, humidity, air filtration), qualified and calibrated equipment, trained and periodically retested personnel, and a documented quality program with corrective-action procedures and annual audits. In practice this typically means classified cleanrooms (commonly ISO 7 or ISO 8) with continuous particulate and microbial monitoring — a level of infrastructure that costs millions of dollars to build and maintain, and that a standard outpatient procedure room does not provide.

Expansion, Bioreactors, and the Passage-Number Ceiling

To generate a clinically useful dose, isolated cells must be grown, or "expanded," through repeated rounds of culture — each round called a passage. Small-scale manufacturing uses flat culture flasks; larger-scale operations increasingly use closed bioreactor systems (hollow-fiber or microcarrier-based stirred tanks) that can expand cells by orders of magnitude in a controlled, contamination-resistant environment. But cells don't stay the same as they divide. A 2019 study in Stem Cells Translational Medicine tracking umbilical cord-derived mesenchymal stromal cells through serial passaging found that significant transcriptome drift didn't appear until around passage 6–9, with cells reliably expandable to clinical dose by passage 4 "before significant transcriptome drift," while replicative senescence — the point where cells stop dividing and lose function — set in variably after passage 10. This is why credible manufacturers cap passage number in their release specifications: beyond a certain point, potency and genomic stability can drift even though the cells still look normal under a microscope. On the encouraging side, closed, automated bioreactor and cell-processing platforms (such as Miltenyi's CliniMACS Prodigy, cited by industry press covering manufacturing trends) are increasingly able to standardize these expansion steps and adjust conditions in real time, which developers say meaningfully reduces batch-to-batch variability and, over time, manufacturing cost — a genuine advance for an industry that has struggled with inconsistency.

Release Testing: The Gate Nothing Is Supposed to Skip

Before a cell lot can be released for clinical use, it's supposed to pass a battery of quality-control assays: sterility testing (checking for bacterial and fungal contamination), mycoplasma testing (a contaminant invisible on standard sterility tests), endotoxin testing, karyotyping (confirming the cells haven't developed chromosomal abnormalities during expansion), identity testing (confirming the cell population expresses the expected surface markers), and potency assays that test whether the cells actually perform their intended biological function, not just that they're alive. This is standard practice at contract testing labs and CDMOs serving licensed cell-therapy manufacturers. It is also the step most commonly cut, minimized, or skipped entirely at clinics manufacturing unapproved products in-house.

Cryopreservation, Storage, and Shipping

Cells intended for banking or shipment are typically frozen with a cryoprotectant (commonly dimethyl sulfoxide, around 5–10% concentration) using controlled-rate freezing to limit ice-crystal damage, then stored in vapor- or liquid-phase nitrogen below roughly -130°C to -150°C to preserve viability long-term. For shipping, specialized cryogenic dry shippers can hold that temperature range for up to about two weeks, enabling global distribution — but temperature excursions during transit remain a real risk that can silently degrade viability and potency before a product ever reaches the patient.

The Regulatory Line: 21 CFR Part 1271 and Why "Registered" Isn't "Approved"

This is the piece clinic marketing rarely explains clearly. Under 21 CFR Part 1271, an HCT/P that is minimally manipulated, used for its normal ("homologous") function, and meets a handful of other narrow criteria can be marketed under the lighter Section 361 pathway — registration and cGTP compliance only, no premarket review. Everything else falls under Section 351 and requires full drug/biologic development, including a Biologics License Application (BLA) — the same pathway used for any other approved biologic. Very few cell products have actually cleared that bar. StemCyte's REGENECYTE, a cord-blood-derived hematopoietic stem cell product, received BLA approval in November 2024 for a narrow, specific use — supporting blood-and-immune-system transplants — making it one of a still-small handful of FDA-licensed cell therapies. Meanwhile, many clinics market cord tissue, adipose SVF, or bone-marrow "stem cell" injections for joint pain, autoimmune disease, or anti-aging under the claim that they're minimally manipulated 361 products exempt from FDA review. Courts have repeatedly disagreed: in a case brought against US Stem Cell Clinic, a federal judge issued a permanent injunction in 2019 after finding the company's adipose-derived stromal vascular fraction product was adulterated and misbranded — i.e., not exempt at all. The International Society for Cell & Gene Therapy (ISCT) has separately warned that many unproven products lack standardized manufacturing controls, are sometimes purchased and resold with no oversight, and are marketed without the preclinical and quality data regulators require.

Bottom Line

Manufacturing quality is not something you can see in a treatment room, which is exactly why it's worth asking about directly. A reasonable set of questions for any clinic or provider: Is this product FDA-licensed (BLA-approved) or is it investigational/unapproved — and if investigational, is it being given under an FDA-authorized IND with IRB oversight? What is the manufacturing site, and is it a GMP facility with third-party inspection, or an in-house lab attached to the clinic? Will they show a certificate of analysis documenting sterility, mycoplasma, identity, and potency testing for the specific lot being used? And is the manufacturing facility FDA-registered — a fact that can often be checked against FDA's public Human Cell and Tissue Establishment Registration database. None of this guarantees a treatment works, but a provider who can't or won't answer these questions is telling you something important about how — and whether — their product was actually manufactured to a verifiable standard.