On this page
- What the trials consistently report
- The four qualifiers that get dropped
- Theoretical risks and what is known about them
- Where the actual harm has happened
- Procedure risk is separate from cell risk
- What good release testing covers
- What would change the picture
- Frequently asked questions
- Requesting a medical evaluation
“Stem cells are safe” is the most repeated sentence in this industry and the least precise. The safety literature supports a narrower claim: that in selected trial patients, receiving a characterised cell preparation, short-term tolerability has generally been good — which is not the same sentence at all.
What the trials consistently report
Across the randomised literature the commonest adverse events are local and transient: pain and swelling at an injection site for a few days, occasionally a brief low-grade fever after an intravenous infusion.
Safety is reported alongside efficacy in the major knee osteoarthritis synthesis — Efficacy and safety of mesenchymal stem cells in knee osteoarthritis — and in the phase III adipose trial — Clinical Efficacy and Safety of the Intra-articular Injection of Autologous Adipose-Derived Mesenchymal Stem Cells.
Outside orthopaedics, safety data comes from trials in diabetic kidney disease — Safety and Preliminary Efficacy of Mesenchymal Stromal Cell (ORBCEL-M) Therapy in Diabetic Kidney Disease — traumatic brain injury — Mesenchymal Stromal Cell Implants for Chronic Motor Deficits After Traumatic Brain Injury — and spinal cord injury — Safety and potential efficacy of expanded mesenchymal stromal cells of bone marrow and umbilical cord.
Taken together that is a reasonably reassuring short-term picture, and it is worth saying so plainly. The problem with the safety claim is not that it is false. It is that it is repeated without the conditions under which it was established.
The four qualifiers that get dropped
Each of these narrows the claim considerably, and each is routinely omitted.
- In selected patients. Trials exclude active malignancy, active infection, significant organ failure and poorly controlled comorbidity.
- Short-term. Follow-up is mostly months. Late and rare events are what short follow-up misses.
- Characterised products. Trial cells are tested. Safety data does not transfer to a preparation nobody has tested.
- Specific routes. A knee injection safety record says nothing about intrathecal administration.
“Safe” without those qualifiers is a marketing claim rather than a finding.
Theoretical risks and what is known about them
Tumour formation. The concern is that cells capable of proliferating might do so inappropriately. MSCs are not pluripotent and this risk is considered low, but cases following unregulated injection into the nervous system are documented.
Ectopic tissue. Cells forming the wrong tissue in the wrong place — bone in soft tissue, for instance. Rare, reported, and more plausible with poorly characterised preparations where nobody has confirmed what the cells actually are.
Immune reactions to donor cells. Uncommon in trial data; the question sharpens with repeat dosing — see autologous versus allogeneic.
Embolism. Cells given intravenously are largely trapped in the lungs on first pass, which raises a theoretical concern more relevant at high doses and with poorly prepared suspensions containing clumped cells.
Where the actual harm has happened
Overwhelmingly outside trial conditions. Contamination of preparations made without supervision. Injection into sites and by routes never studied. Products sold for conditions with no evidence at all.
The medical tourism literature has examined the broader pattern — Patient care without borders: a systematic review of medical and surgical tourism.
The distinction matters because the trial safety record is used to reassure about a sector the trials never covered — see adverse events.
Procedure risk is separate from cell risk
Septic arthritis after a joint injection is uncommon and serious. Donor-site complications follow any harvest. Intravenous access carries its own small risks. Intrathecal injection carries considerably more.
None of these are risks of the cells. All of them are risks of the treatment, and they are what a consent conversation should actually cover.
For anyone travelling, there is a further layer: a complication appearing after you fly home is managed by doctors who did not perform the procedure, do not know the product, and have no relationship with the clinic that supplied it — see international patients.
A clinic describing cell safety without describing procedure risk has answered a different question from the one you asked.
What good release testing covers
Sterility — bacterial and fungal culture. Endotoxin. Mycoplasma. Viability, measuring what proportion of cells are alive at the point of use. Identity, confirming the cells are what they are supposed to be.
These are the checks that prevent the harms that have actually occurred, and they are the most useful thing to ask any clinic about.
Ask also who performs them — an in-house check and an independent laboratory report are different levels of assurance, and the answer is usually given readily by clinics that use the second.
A clinic that can name its release criteria is describing a process. One that talks instead about the quality of its equipment is describing a room — see MSC therapy for what those steps involve.
What would change the picture
Registries following treated patients for years rather than months. Some exist; participation is voluntary and incomplete, which means the denominator is unknown and rare events cannot be counted properly even in principle.
Mandatory adverse event reporting in the commercial sector, which largely does not happen.
Until then, the honest position is short-term tolerability looks good, long-term safety is genuinely unknown, and the harms on record are mostly preventable ones — see the research library.
Preventable is the operative word. The documented harms in this sector were not surprising biological events. They were contamination, untested routes and treatment of people who should have been excluded, and each of those is caught by a process rather than by luck.
Frequently asked questions
Is stem cell therapy safe?
In trial conditions — selected patients, characterised products, months of follow-up — short-term tolerability has generally been good. That is a much narrower claim than “safe”, and every one of those qualifiers is doing work.
What are the common side effects?
Transient pain and swelling at an injection site, and sometimes a brief low-grade fever in the hours after an intravenous infusion. Both usually settle within a few days without treatment.
Could it cause cancer?
MSCs are not pluripotent and the risk is considered low. Cases following unregulated injection into the nervous system are documented, which is a different situation.
What about long-term safety?
Genuinely unknown. Most follow-up is months, and rare or late events are exactly what that misses.
What should I ask a clinic about safety?
Its release criteria — sterility, endotoxin, mycoplasma, viability and identity testing. Those prevent the harms that have actually occurred.
Does the trial safety data cover clinic treatment?
Not really. Trials use selected patients and characterised products, and the documented harms have mostly occurred outside those conditions.
Requesting a medical evaluation
Nothing on this page establishes whether any treatment is appropriate for you. That needs your history, your imaging and your current medications read by a clinician. If you would like that review, send your records and we will tell you honestly whether there is anything worth discussing — including when the answer is no.
This page is general information, not medical advice, and does not create a doctor–patient relationship. Regenerative treatments discussed here are in most cases investigational. Discuss any treatment with a clinician who knows your history.