Condition · Knee pain
Stem cell therapy for knee pain: what may help, what to question.
Most people do not start with a stem-cell question. They start with a knee that makes stairs feel older than they are, a doctor mentioning arthritis, or a clinic promising a way to delay surgery. PRP, BMAC, exosomes, and stem-cell injections sit in that hopeful middle: interesting biology, mixed evidence, and a lot of marketing. The useful question is not does it regenerate my knee? It is what might it realistically help, for which knee problem, and under whose oversight?
Written byJimmy L Wu
Sources we readFDA · EMA · PMDA · MFDS · AAOS · ISSCR · Cochrane
Research checked2026-05-14
What patients are really asking.
The big question is usually, will this regrow my knee? Right now, human studies do not support that promise for arthritis injections, no matter what is in the syringe. The smaller questions are the useful ones: can it reduce pain, calm the inflammation behind flare-ups, make stairs and walking easier, or buy time before a joint replacement that may still be the right answer eventually?
For a few products, there are real human trials behind one or two of those goals. For most, the marketing runs well ahead of what the studies actually show. That is the gap this page tries to close.
We read the field by product and by knee diagnosis, in plain English, with every regulatory and clinical claim tied to a source at the bottom. No clinic recommendations. No medical details collected. The goal is to make the conversation with your doctor a sharper one.
Not magic, not nothing either.
Regenerative medicine for the knee is a real, active field. The cell biology is interesting. A handful of products are approved abroad for very specific cartilage-repair problems. Human trials keep coming. The field is not made up, and it is not standing still — that is why patients are paying attention.
It is also not all the same thing. The most useful way to read a clinic page or a study is to keep two questions apart.
The realistic question. Pain, inflammation, function, possibly delaying surgery. These are the things the better human trials actually measure, and where there is sometimes something real worth a closer look.
The much harder claim. Regrowing cartilage, reversing bone-on-bone arthritis, or showing cartilage change on a follow-up scan. No injectable regenerative product in current clinical use has been shown to do that reliably in human osteoarthritis. Bone-on-bone arthritis is not reversed by any approved injection. Clinics should not sell that as a proven result. The chart below puts those two kinds of question side by side.
Promise vs reality.
Four questions a patient might ask a clinic, sorted by how answerable they currently are.
Pain and inflammation
A fair question to ask.
Several products have human studies aimed at pain and the inflammatory signals behind it. Effects are often modest and short-term, and trials don't always agree — but there is something real to investigate here.
Function and activity
A fair question, with a smaller effect.
Function scores often move with pain scores, but the size of the change is usually smaller than what patients consistently feel as a meaningful improvement.
Delaying surgery
An open question.
Some cohort studies hint that selected patients defer joint replacement after these treatments. The evidence is suggestive rather than proven, and surgery delay is one of the more honest and important goals still being studied.
Cartilage regrowth
Not a fair question to ask of these injections.
No injectable regenerative product in current clinical use has been shown to reliably regrow articular cartilage in human osteoarthritis. Bone-on-bone arthritis is not reversed by any approved injection. The biology is still being studied — clinics should not sell that as a proven result.
The bars are illustrative of how strong the evidence is, not of a single patient’s expected benefit. Sources for each row are cited in the per-product cards and footnotes below.
What clinics mean by “stem cells for knees.”
Most of what clinics call “stem cells” for knees is not purified stem cells. Some products contain no cells at all. Here is what each label usually means.
- PRP
- A tube of your own blood spun in a centrifuge to concentrate platelets and the growth factors they carry. Not a stem-cell product. The centrifuges and tubes are FDA-cleared as medical devices for processing blood; that clearance covers the equipment, not the claim that PRP works for knee arthritis.
- BMAC
- Bone marrow aspirate concentrate. Marrow drawn from your pelvis and concentrated in a centrifuge. A mixed product from your own body, containing some stem and signaling cells along with platelets and other components — not a purified stem-cell product. Used widely in U.S. orthopedic practice, but not FDA-approved as a drug for knee arthritis.
- Adipose / SVF
- A cell mixture isolated from your own fat tissue. How the fat is processed matters a lot. Enzymatic processing is treated as drug-territory by the FDA, and no SVF product is approved in the U.S. for knee arthritis. Mechanically processed micro-fragmented adipose is a related but distinct category.
- Culture-expanded MSCs
- Mesenchymal stromal cells grown in a laboratory over weeks to reach therapeutic doses. The source can be bone marrow, fat, or perinatal tissues; the cells can be your own or from a donor. Growing them in a lab moves the product into U.S. drug-approval territory, and no culture-expanded MSC product is FDA-approved for knee arthritis. South Korea has approved one — for surgical implantation into cartilage defects, not for routine knee injection.
- Exosomes
- Tiny signaling vesicles, about the size of a virus, released by cells as paracrine cargo. A genuinely interesting area of cell biology. In the United States there is no FDA-approved exosome product for any clinical use, and the FDA has issued a public safety notification specifically about unapproved exosome injections.
- Amniotic / Wharton’s jelly
- Donor-derived products processed from amniotic fluid, amniotic membrane, umbilical cord tissue, or Wharton’s jelly. Often marketed as “young stem cells”; most processed injectable forms contain few or no viable stem cells by the time they reach the patient. The FDA’s view is that most marketed umbilical and Wharton’s jelly products injected for knee arthritis fall outside the limited tissue-product exception, and warning letters have been issued.
- Cord blood
- Worth a clarifying word: FDA-approved cord-blood products in the United States are blood-forming stem cells used in transplants for leukemias, lymphomas, and certain inherited disorders — not knee injections.5 When a clinic uses “cord blood” in a knee context, they are almost always describing a perinatal MSC product or tissue allograft, which is a different category with the regulatory picture above.
What each option is really being sold as.
Six treatments, the realistic thing each is being investigated for, the part to be careful about, and one question worth asking in the consult room.
PRP
What it is
Your own blood, spun in a centrifuge to concentrate platelets and the growth factors they carry. The most familiar option here, and the least “stem-cell.”
The realistic hope
Pain relief and modest function improvement for some patients with mild-to-moderate knee arthritis.
What to be careful about
The evidence is genuinely mixed: one large placebo-controlled trial was negative, while some reviews still find a positive signal. Major guideline groups do not all read the data the same way.6,7,8,9 The real question is whether the clinic’s exact PRP recipe — concentration, whether white blood cells are included, how many injections — matches the studies it cites.
Question to ask
Which PRP protocol do you use, and what human study does it most closely match?
BMAC
What it is
Bone marrow drawn from your pelvis and concentrated. A mixed product from your own body — some signaling cells and growth factors, not a purified stem-cell product.
The realistic hope
Pain and function improvement for some patients with mild-to-moderate knee arthritis. The biology is more interesting than pure platelet PRP.
What to be careful about
Pain and function scores tend to move in the right direction in studies, but the difference vs. simpler injections is usually not large enough for patients to feel reliably.10 The main U.S. orthopedic society has reviewed the literature without taking a position for or against.11
Question to ask
What is actually in the concentrate, and what knee diagnosis was studied with a similar protocol?
Adipose / SVF
What it is
A cell mixture isolated from your own fat tissue.
The realistic hope
Symptom improvement for some patients with mild-to-moderate arthritis.
What to be careful about
How the fat is processed matters a lot. Enzymatically processed SVF is treated by the FDA as a drug — and none is FDA-approved for knee arthritis.2 Most published studies are small and not blinded, which makes positive results harder to read.12
Question to ask
How is the fat processed, and what local regulatory or hospital-review pathway allows that processing for knee treatment?
Culture-expanded MSCs
What it is
Stromal cells grown in a laboratory to therapeutic doses, from bone marrow, fat, or perinatal tissue. May be your own cells, or a donor’s.
The realistic hope
Calming inflammation and improving pain and function in selected patients with mild-to-moderate arthritis. One of the more active areas of human-trial work in the field.
What to be careful about
The products vary widely — cell source, dose, and whether the cells come from you or a donor all change the picture. None is FDA-approved for knee arthritis. In the U.S., a clinic offering this outside a registered clinical trial is doing so without FDA oversight, and the main international stem-cell research society has cautioned against that kind of marketing.13,1,14
Question to ask
Are these cells expanded in a lab, who manufactures them, and what regulator or clinical-trial pathway oversees their use for this knee condition?
Exosomes
What it is
Tiny signaling particles released by cells. About the size of a virus.
The realistic hope
A research-stage idea that the signaling cargo of cells could ease inflammation without injecting the cells themselves. Genuinely interesting biology.
What to be careful about
Human knee-arthritis data is small and very early. The FDA has not approved any exosome product for any use, and has reported serious adverse events from unapproved exosome injections.3 Commercial claims are well ahead of the human evidence.
Question to ask
Who manufactures the exosome product, what quality testing is provided, and what human knee data supports this exact use?
Amniotic / Wharton's jelly products
What it is
Donor-derived products processed from amniotic fluid, amniotic membrane, umbilical cord tissue, or Wharton’s jelly. Not your own cells.
The realistic hope
A pain-relief option in some early trials.
What to be careful about
Product identity varies a lot by manufacturer, and the FDA has issued warning letters to umbilical and Wharton’s jelly companies marketing these for orthopedic injection.4 One blinded human trial against a standard steroid injection found the two were roughly comparable on knee outcomes.16
Question to ask
Which manufacturer’s product is this, what is still biologically active in it, and what regulator or hospital pathway allows it for knee treatment?
Each card cites the trial, review, or regulator behind its “careful” line in the footnotes below. Active human trials for each product can be looked up in the public registry.15
The knee diagnosis matters more than the cell type.
“Knee pain” is a complaint, not a diagnosis. A small contained cartilage chip in an otherwise-healthy knee is a different problem from long-standing wear-and-tear arthritis. A torn meniscus is different again. So is inflammatory arthritis. A study in one knee problem does not automatically apply to another. A trial in mild-to-moderate arthritis says little about severe bone-on-bone disease; a study in cartilage defects says little about general osteoarthritis.
- Mild-to-moderate osteoarthritis. The population most trials enroll. Pain and function effects of PRP, BMAC, SVF, and culture-expanded MSCs have been studied here. Durability beyond a year or two is largely an open question.
- Severe / bone-on-bone osteoarthritis. Trials specifically in this group are scarcer. None of the injection products on this page has been shown to reverse structural arthritis at this stage. A conversation about timing of knee replacement belongs here alongside any injection question.
- Meniscus injury. Repair, partial meniscectomy, or conservative care is the standard frame. Regenerative add-ons (PRP at the time of repair, for example) are an active research area; results vary by tear pattern and the procedure being augmented.
- Focal cartilage defect in an otherwise-healthy knee. A specific diagnosis with its own surgical literature — microfracture, autologous chondrocyte implantation, osteochondral grafting. The EU and Japan have approved cultured-chondrocyte products for this; South Korea has approved an umbilical-cord-blood-MSC product for cartilage defects in osteoarthritis (see §8). They are surgical implants, not injections, and the indication is the cartilage defect, not osteoarthritis.
- Ligament or tendon injury around the knee. Different evidence base. What holds in joint-injection trials for arthritis does not automatically transfer.
- Inflammatory arthritis (rheumatoid, psoriatic). Systemic diseases managed primarily with disease-modifying medications and biologics, with a rheumatologist. Local regenerative injections are not the conversation here; treating the systemic disease is.
- Post-surgical knee pain. A separate workup — hardware, alignment, infection, complex regional pain syndrome — before an injection conversation makes sense.
One useful thing to bring to a consult, if your clinic has it: the x-ray severity grade or imaging diagnosis from your most recent knee scan.
What abroad changes — and what it does not.
Some of the most interesting work in this field is happening outside the United States. The EU, Japan, and South Korea have each approved specific cell-therapy products for narrowly defined cartilage-repair problems. Those approvals matter: serious regulators have looked carefully at specific products, and the biology produced clinical results worth licensing.
They are also narrow. Each one is for a specific product, in a specific procedure, for a specific kind of knee damage — usually a small, contained cartilage defect, surgically implanted in an otherwise reasonably healthy joint. None of them is a blanket endorsement of “stem cell knee injections” as a category.
What approvals abroad show
The field is real and moving. Serious regulators have looked carefully at specific products and licensed them for narrowly defined cartilage-repair problems. Some of the cell biology has produced clinical results worth licensing.
What they do not prove
These are narrow surgical-product approvals for specific cartilage defects — not general validation of “stem cell knee injections” as sold at most cash-pay clinics. Approval abroad does not transfer to a different product, a different procedure, or your specific knee.
- European Union — Spherox. An autologous cultured-chondrocyte product, authorized by the European Medicines Agency for the repair of contained cartilage defects of the femoral condyle and patella of the knee. It is surgically implanted into the defect, not injected into the joint. The EMA assessment report explicitly states the product must not be used in patients with primary generalized osteoarthritis or advanced osteoarthritis of the knee.17
- Japan — JACC. Autologous cultured chondrocytes from Japan Tissue Engineering. First approved in Japan in 2012 for traumatic cartilage defects and a related condition called osteochondritis dissecans, and more recently extended to a knee osteoarthritis indication, with national insurance reimbursement in 2026. JACC is implanted surgically into the defect site, not injected. Approved in Japan only.18
- South Korea — Cartistem. An allogeneic umbilical-cord-blood-derived MSC product. Licensed by the Korean drug regulator in 2012 for the treatment of cartilage defects in knee osteoarthritis. Like Spherox and JACC, it is surgically implanted into the defect site during an arthroscopic procedure, not injected into the joint space.19
- Beyond these three. Some countries — including Thailand, Mexico, Panama, and others — offer regenerative knee procedures outside the regulatory pathways above. CellDecide is not making a sourced regulatory claim about any of those jurisdictions on this page. Local permission to offer a procedure is not, by itself, evidence of benefit for a specific patient with a specific knee.
Two readings, both true
FDA non-approval in the United States does not make the underlying biology worthless; it means a specific product, for a specific use, has not cleared the U.S. evidence threshold. Approval abroad is meaningful — but it does not transfer to a different product, a different procedure, or a clinic that is not the one running the licensed protocol. Both readings are accurate, and both belong on the page.
How to read the regulatory map
In the United States, FDA status is the main approval question. Abroad, the better question is broader: which local regulator, hospital system, ethics board, trial registry, or ministry pathway allows the treatment? Wherever you are, the question that travels with you is the same — who reviewed this exact product, for this exact knee problem, in this exact setting?
That framing keeps optimism honest. The interesting part of the field is real; the hard part is matching the right claim to the right knee.
Before you pay: three checks.
Before any deposit changes hands — at home or abroad — three questions cover most of what should make a careful patient pause. A clinic doing its job will answer each one in a sentence. If the answers all drift into “every patient is different,” that drift is itself an answer.
1
What exactly is being offered?
“Stem cells for the knee” is not a product, and “knee pain” is not a diagnosis. Get the cell type and the exact knee diagnosis in writing.
- What exactly is being injected — blood, bone marrow, fat, lab-expanded cells, exosomes, or a perinatal product? Is it yours or a donor’s?
- What knee problem is this meant to treat — mild-to-moderate arthritis, severe arthritis, a cartilage defect, a meniscus injury, or something else?
- Is this exact product authorized or studied for that exact knee problem?
- Who oversees this where it is offered — an FDA pathway, a local regulator, a hospital system, an ethics board, or a registered clinical trial?
2
What claim is the clinic making?
Honest claims here are about pain, inflammation, function, and maybe delaying surgery. If the conversation reaches further than that, slow down.
- “Stem cells will regrow your cartilage.”
- “This will reverse bone-on-bone arthritis.”
- “Exosomes are FDA-approved for knee pain.”
- “There is no risk because the cells are your own.”
- The same product is offered for knees, autism, COPD, fertility, and anti-aging.
3
What happens after the injection?
Most of these treatments are paid out of pocket, and complications that show up weeks later are usually managed at home — often by a clinician who did not deliver the procedure.
- What is the follow-up plan at 30, 90, and 180 days, in writing?
- Who handles a complication after you have travelled home?
- What is the total cost — including travel, lodging, repeat doses, and missed work — not just the procedure fee?
- What happens, in writing, if it does not work?
For the longer versions of these three checks, see 20 questions to ask any stem cell clinic before you pay, stem cell clinic red flags, and the stem cell therapy cost estimator for total landed cost. How CellDecide handles affiliate-related disclosures lives in disclosures.
Common questions.
Do stem cells regrow knee cartilage?
No injectable regenerative product in current clinical use has been shown to reliably regrow articular cartilage in human osteoarthritis. Specific surgical procedures exist for focal cartilage defects — autologous chondrocyte implantation in Europe and Japan, umbilical-cord-blood-MSC implantation in Korea — and are approved in their jurisdictions for defect repair, not for treating osteoarthritis in general.171819
Is PRP the same as stem cell therapy?
No. PRP is concentrated platelets and plasma; it contains no stem cells. PRP is sometimes marketed as a “stem cell injection,” which is inaccurate to what is actually in the syringe.
Is BMAC FDA-approved for knee arthritis?
No. BMAC is used widely in U.S. orthopedic practice under a “same surgical procedure” pathway, but it is not an FDA-approved drug for knee arthritis. The 2023 AAOS Technology Overview of BMAC summarized the literature without issuing a guideline recommendation for or against.11
Are exosomes approved for knee pain?
No. As of 2026, there are no FDA-approved exosome products for any clinical use — knee or otherwise. The FDA has issued public safety notifications, and reports include serious adverse events from unapproved exosome injections.3
Can regenerative medicine delay a knee replacement?
Maybe, in selected patients with mild-to-moderate disease. This is one of the more honest open questions in the field. Some follow-up studies report patients deferring surgery; larger controlled trials that track time to replacement as the main outcome are scarcer. The right reading is “an open question worth a careful conversation,” not “a proven way to avoid surgery.”
Is it better to get stem cell treatment abroad?
It depends on what is being offered, where, and for what diagnosis. Specific cell-therapy products are approved in the EU, Japan, and Korea for focal cartilage defects, surgically implanted, not for general osteoarthritis injections. Local permission to offer a procedure is not, by itself, evidence of benefit for a specific knee. Travel also moves complication management back home, where the patient’s home-country physician may not have the chart.