The document called the Certificate of Analysis.
Every cGMP-released batch of cells ships with a one-page document called a Certificate of Analysis (CoA). The CoA lists every assay run on the batch, the specification each had to meet, and the actual measured result. If any assay is out of spec, the batch is quarantined and rejected. It does not reach patients.
The CoA is not a marketing document. It's a release record, signed by a Quality Assurance officer with their name and date on the bottom. It is the single most important piece of paper in stem cell manufacturing, and it is the thing every prospective patient should ask to see.
This page walks through every assay on a complete CoA, in the order it would appear.
Sterility, nothing else in the vial.
The first job of the testing panel is to prove that the only living things in the vial are MSCs. Anything else: bacteria, fungi, mycoplasma, viral contaminants, would be injected directly into a patient, often into the bloodstream. The consequences range from local infection to systemic sepsis, with a real mortality rate. cGMP exists primarily to prevent this.
Rapid result gates the ship. 14-day test confirms.
A modern cGMP lab runs BacT/ALERT rapid sterility in parallel with USP <71>. The rapid result is back in time to release the batch; the longer culture continues in parallel and triggers a recall if it ever fails.
Bacteria + fungi
Detected by 14-day USP <71> culture and 7-day rapid systems.
Mycoplasma
The most common cell-culture contaminant. Invisible to USP <71>. PCR-detected.
Endotoxin
Bacterial-cell-wall fragment. Detected by LAL assay even when no living bacteria remain.
Viral panel
Donor pre-screen: HIV, HBV, HCV, HTLV, syphilis, CMV. PCR + serology.
Identity. that they are MSCs.
A cell count tells you how many cells. An identity assay tells you what kind of cells. The International Society for Cellular Therapy (ISCT) published minimal criteria for what counts as a mesenchymal stem cell in 2006; these are the global standard. Every cGMP MSC product is released against them.
Three markers must be there. Three must not.
Each plot is a flow-cytometry histogram: x-axis is fluorescence intensity, y-axis is cell count. A positive marker pushes the peak to the right (≥95% positive). A negative marker keeps it on the left (≤2%).
Positive markers
Negative markers
The ISCT criteria are not just "has these markers." They are "has these AND lacks these." A cell with the right positive markers but the wrong negative markers is contaminated. it's not a pure MSC product. A vial with 15% CD45 is contaminated with white blood cells; with 10% CD34, with hematopoietic stem cells. Neither is a pure MSC product. This is why the "100M cells" headline is meaningless without a flow panel.
Viability. that the cells are still alive.
Identity proves the cells are MSCs. Viability proves they're alive and metabolically active when they reach the patient. Therapeutic effect depends on living cells: dead cells don't signal, don't release a useful secretome, and in fact can produce a pro-inflammatory response, the opposite of what's intended. Viability is measured by 7-AAD flow cytometry post-thaw, on the actual lot being released.
The physician team's laboratory holds every batch to a post-thaw viability standard of 95% or above before any treatment proceeds, well above the regulatory minimum for cell viability. Batches that do not meet this threshold are not used. They are discarded, not "released with a note." Viability is a release criterion, not a marketing claim. The number on the Certificate of Analysis is the number from the bag the patient receives, not a number from before freezing.
Ask: "Is viability tested on the day of treatment, post-thaw, on the same lot I'm receiving, and what is the release floor?" The answer should be yes, post-thaw, and a specific number the lab commits to. A clinic quoting only a manufacturing-time viability number is quoting from before freezing, which is not the number that matters.
Potency. that they still function.
Identity proves the cells are MSCs. Potency proves they still do the things MSCs are supposed to do. Cells can be alive, the right phenotype, and yet senescent: they no longer divide, no longer secrete the cytokines that drive immunomodulation, no longer differentiate. Potency assays catch this.
Five functional tests against threshold.
The threshold is the minimum the cells must demonstrate to be considered functional. Bars below show measured potency for a representative lot; the dark line marks the release spec.
Sterility and identity are binary. Potency is graded, a cell can be 30% as potent as a fresh-from-tissue cell and still pass an identity panel. The newest FDA guidance for cell therapies (2023) requires sponsors to define and validate a potency assay relevant to the proposed mechanism of action. Operations selling MSCs with no potency data are operating below the regulatory standard required for clinical trials.
"Our cells are viable and pass identity" is not a complete answer. Both can be true while the cells are functionally dead, alive but no longer secreting the cytokines that make MSC therapy work. Ask: "What's the potency spec for this batch, and what was the measured result?"
Residuals. what's in the vial that isn't cells.
Manufacturing MSCs requires reagents: growth media, antibiotics, dissociation enzymes, cryoprotectants, serum. By release, these should be reduced to trace amounts or removed entirely. Residual testing quantifies whatever remains.
What gets quantified, and limited.
Not every test on every batch.
Some assays are one-time qualifications on the Master Cell Bank that get inherited. Others run on every working bank. The most critical run on every final-product lot before release.
- Donor eligibility
- HIV / HBV / HCV
- HTLV / syphilis
- Medical history
- Tumorigenicity
- Karyotype
- Identity (full)
- Sterility
- Identity (flow)
- Sterility
- Mycoplasma
- Morphology
- Sterility (rapid)
- Endotoxin
- Viability (post-thaw)
- Potency · Residuals
MSCs grown through many passages can accumulate chromosomal changes, most don't, but some can. A karyotype confirms cells released for clinical use have a normal chromosome complement. Same-day, non-expanded preps don't need it because the cells haven't been through enough divisions to develop abnormalities, but they also haven't been characterized at all.
The Certificate of Analysis. what it should contain.
A complete CoA is one page (sometimes two), with a header, the results table, and a signed-and-dated release line at the bottom. Below: a sample Certificate of Analysis for a representative cGMP MSC batch.
Certificate of Analysis
Page 1 of 1
What a "results brochure" is not
Some clinics show patients a glossy summary with phrases like "our cells are tested for sterility and viability." This is not a Certificate of Analysis. The CoA is a regulatory document tied to a specific lot number, signed and dated, with measured values. If a clinic offers you a marketing brochure when you ask for the CoA, they are showing you the wrong document, and likely do not have access to the right one.
"Does your clinic produce a CoA for each lot and make it available to patients on request?" A real cGMP operation produces and retains a CoA for every batch by SOP. A clinic that hand-waves the question, or has no CoA at all, is operating outside cGMP.
Fifteen assays, one document, signed and dated.
A real cGMP-released cell product comes with a Certificate of Analysis. Sterility, mycoplasma, endotoxin, identity by flow cytometry, viability, potency, residuals, karyotype, each with a spec and a measured value. One out-of-spec result and the batch is rejected.
You don't need to read it like a scientist. You need to know it exists, ask for it before you commit, and treat its absence as the answer it is.
Ask for the CoA
Every batch generates one. Your dose came from a specific lot. The CoA is available on request through your case coordinator.
Look for QA signature
Named person, role, date. Not a marketing summary, not a glossy. A regulatory document with a name and a stamp on it.
Check the lot match
The lot on the CoA should match the lot on the vial you're receiving. Not "a representative batch". your batch.